This dissertation has been submitted by a student. This is not an example of the work written by our professional dissertation writers.

CHAPTER 1INTRODUCTION

What is an Office?

Offices are workspaces designed for regular use to achieve personal, group or organizational goals through the accomplishment of tasks. Sanders and McCormick, (2002) go on to state that these tasks can be grouped into cognitive, physical, social or procedural tasks. The office provides a location for contact and could also be a repository for tools, information and other resources required to meet business objectives. It is also a business resource, this a point most people fail to understand thus, the failure to properly design and evaluate work spaces.

The work place or office is one of the places the modern man spends the bulk of his waking hours. Sanders and McCormick, (2002) say almost half of ones waking hours are spent in and around the office. This would provide explanation for research efforts into the design and utilization of offices.

According to Myerson and Ross (2003) the office grew out of the factory and then followed the trend of bureaucratization of industry. Thus, offices have been viewed differently by users and companies. Some view it as an address, others as a necessary evil but to others it is considered to be an asset. Bjerrum and Bødker (2003) noted that the design of an office was mostly considered as a cost and done to support quiet work and also show peoples status. While the purpose of the “New office” is to be that of attracting and retaining staff as well as to revolutionalize corporate culture.

Work places or offices have been described variously as; conventional, traditional, and closed or open plan offices. Some, group them as large or small, landscaped etc. (Sanders and McCormick, 2002). The general descriptions of workspaces fall under the categories of open plan and cellular offices and this is based on the architectural and functional features of the work spaces (Duffy, Laing and Crisp, 1992).

Other descriptions and categorization of offices include the hive; which is suited to individual processes. The den; suited to group processes. The cell; designed for concentrated study while the club supports transactional knowledge (Sailer, Budgen, Lonsdale, Turner and Penn, 2009). There are other descriptions of office types for example, Myerson and Ross (2003) from an architectural point of view, showed that views of property and space as related to the office environment have been evolving and as such, they identify four thematic categories of offices namely: narrative which presents the “office as a brand experience.” Nodal where the “office as knowledge connector.” The neighborly theme sees the “office as a social landscape” and lastly nomadic “office as distributed work space” these grouping reflect more of necessity and corporate culture not necessarily a collection of generally practically replicable models.

In another categorization of offices by Myerson and Ross (2006) is based on the fact that the offices evolved to suit knowledge workers, as such, the categories match each of the four “realms” of knowledge work namely: the academy “is likened to the corporate realm which is a more collegiate and collaborative approach to work”, guild “the professional realm in essence a professional cluster of peers sharing a skill or specialization”, agora “the public realm where the corporation is open to the city or the market place” and the lodge “ the domestic or private realm more of the live and work setting”. In the general scheme of things offices are still broadly classified into open plan and private or cellular offices all other forms are variations of the two.

Cellular Offices

Also called closed offices, this type of offices are the traditional or conventional offices which are usually closed and private workspaces (Maher and von Hippel, 2005) i.e. they are designed with floor to ceiling walls, a door and dimensioned for a single user. This type of office is also called a cell-office and can be a shared room office, used by 2-3 persons (Danielsson, 2008). This has been the generally accepted, traditional or popular understanding of the place called an office.

Open Plan Offices

These are found to be a common workspace shared by a group of employees. The original concept of the open plan office has continued to evolve, but it is the absence of floor-to-ceiling walls that is said to be the primary characteristic of open-plan offices. The arrangements of office furniture, partitions, screens, office equipment, or plants mark out individual and functional work areas (Valesny and Farace, 1987).

One of the strengths of the open plan office according to Bjerrum and Bødker (2003) is the openness and flexibility allowing one to move to where things are happening and allowing for “overhearing and over-seeing” (p. 207) thus enhancing peripheral participation.

Other types of the open plan office include the bull pen office, action offices, landscaped offices (Sanders and McCormick, 2002). In the bull pen offices, the work desks are arranged in neat row as far as the eyes can see.

In reality, most firms have a mix of office typologies ranging from cellular units designed for a single user to a small room office shared by a few people then the spaces shared with a large group with or without specifically assigned work places and with varying measures of visual and audio privacy.

Recent Developments in Open Plan Offices

It is safe to argue that, the open plan office has become increasingly popular (de Korte, Kuijt-Evers and Vink, 2007; Ding, 2008; Oldham and Brass, 1979; Pejtersen, Allermann, Kristensen and Poulsen, 2006 etc.) and several reasons could be advanced to explain the widespread adoption and use of the open plan offices and its variations.

There is also a move to wards a reduction in open plan office workspaces especially in the United States of America due to the understanding that smaller workstations are cheaper to maintain (Dykes, 2011) this according to Veitch, Charles, Farley and Newsham (2007) is because there is a failure in understanding the full value of the physical office environment and related issues in open plan offices in particular.

Advantages of open plan offices

Searches through literature (Danielsson, 2008; Oldham and Brass, 1979; Pan and Micheal, 2007; Roper and Juneja, 2008; Valesny and Farace, 1987 etc.) present the following as reasons for the adoption of open plan offices. They include;

  1. Reduction in office space and cost decline: The price of real estate is predicated on the area rented and utilized. With organizations using rental spaces, it is cheaper to use the rented floor or floors as open plan offices. In most cases, the cost of partitioning is saved if an open plan set up is deployed fully or partly.
  2. Flexibility for organizational changes: The open plan office lends itself to easy restructuring of work areas. In most cases, it is easier to fit in one more members of staff (Sanders and McCormick, 2002).
  3. More efficient work flow and communication: Some jobs require continuous team work, face to face interaction and a relatively high level of routine procedures. For such work groups, the open plan office or variations thereof are usually recommended and deployed. The enhancement of some level of peripheral participation is one of the strengths of the open plan office.
  4. Possible enhancement of social facilitation: The enhancement of collaboration i.e. the fostering of a team spirit, where, work teams or task forces are close to one another and can quickly form a huddle to sort out problems without resorting to information technology provisions like the intercom, emails, phones, video conferencing or even the walk up to another office. Oldham and Brass, (1979) specifically examined interpersonal issues that included; intradepartmental and interdepartmental interaction, friendship opportunities, noting that supervisor and co worker feed back could be improved.
  5. Ease of supervision: There is an ease of supervision, in that, a look over the landscape of the office can give an idea as to who is present and what each member of staff is doing.

Limitations of open plan offices.

Regarding the limitations of open-plan office designs, Maher and von Hippel (2005) rightly point out the fact that in open plan office layouts “distractions and overstimulation are intrinsically linked to the design.” These issues have consistently been themain down sides of open plan offices and some of them include:

  1. Increased workplace noise (Pan and Michael, 2007).
  2. Increased disturbances and distractions.
  3. Increased feelings of crowding and loss of privacy.
  4. There is a reduction in autonomy and task identity and a reduction in supervisor and co - worker feedback in certain cases (Oldham and Brass, 1979).

One point of agreement in open plan office research is that there is a generally low level of perceived privacy in open plan offices, as interruptions and distractions of the visual and acoustic kind occur frequently in open plan offices. (Pejtersen et al. 2006; Roper and Juneja, 2008)

Furthermore, researchers have observed that these negative outcomes resulting from the adoption of the open plan office design tends to result in dissatisfaction with work and the workplace thus, reducing functional efficiency, decreasing performance, especially, for non routine tasks and also, reduced feedback from supervisors due to some complexity with the freedom of communication (de Korte et al. 2007; Pejtersen et al. 2006; Sundstrom et al. 1982; Vischer, 2007 ). This understanding has led some organizations to begin returning to the traditional private offices i.e. with floor to ceiling partitions assigned to an individual (Roper and Juneja 2008).

Evolving nature of office work

Also worthy of note, is the evolution of work patterns. An increasingly large number of persons work mainly at or from home and visit the office sparingly. This has given rise to the several types of offices one of which is the flex-office, which is dimensioned for less than 70% of the total company staff to be in at the same time. Another design is the combi-office; where a member of staff is not assigned to a specific desk but sharing of common facilities provides the spatial definition of such an individuals work space i.e. the task and personnel at hand may determine the sitting arrangement of persons in the office (Danielsson, 2008).

Thesis Overview

The thesis is organized in to 5 chapters; Chapter one provides an introduction the concept of an office, its major types and variations. It then focuses on the open plan office and then highlights the strengths and limitations of the open plan office.

Chapter two provides a literature review of the concept of privacy perception; it reviews the perceived benefits of privacy and then traces the expectation that privacy perception could be influenced by culture. Significant studies related to dissatisfaction with open plan offices are examined for possible links to culture. The discussion then moves to culture, its definition and then the attempts made in the classification of culture. The Hofstede paradigm is then discussed and some studies employing the paradigm are reviewed. The research motivation and hypotheses are presented.

Chapter three discusses the methodology of the study, the survey method, issues noted and the challenge expected. The source and design of the questionnaire was presented and the analysis methods proposed. The statistical analysis tool was briefly introduced.

Chapter four shows the procedure of the survey, documents the responses received, analyzed the data collected from the general information part of the questionnaire and then chronicles the statistical analysis of the second part of the questionnaire designed to elicit privacy perception in open plan office environments.

Chapter five provides a discussion of the results obtained in chapter five and then presents the limitations of the current study while providing directions for further work.

CHAPTER 2LITERATURE REVIEW

This part of the thesis discusses the links between privacy perceptions and culture. It also includes definitions and explanations of some related terms. Lastly, it includes a presentation of some ideas relevant to the work and results of related studies.

The chapter concludes with the research question, research hypothesis and the motivation for the study.

Privacy Perception

In order to facilitate a better understanding, the term “privacy” is defined firstly then the concept called “perception". According to the Merriam-Webster online dictionary (2011), privacy is said to be “the quality or state of being apart from company or observation”. Wikipedia defines perception as “the process of attaining awareness or understanding of sensory information.” It goes on to say “what one perceives is a result of interplays between past experiences, including one’s culture, and the interpretation of the perceived.”

Privacy is a very difficult concept or construct to define not to talk of evaluating, it has commanded interest from the fields of anthropology, architecture, cultural geography, environmental design, ethology, history, law, philosophy, and sociology, as well as branches psychology such as; clinical, counseling, developmental, educational, environmental and social psychology (Newell, 1995; 1998).

Newell (1995) in her extensive review of the concept of privacy divided the perspectives of privacy into, people centered, place centered and the person-environment or the person-place interaction with the primary interest on the place, people or equally on the person and place and or with the interaction itself. Leino-Kilpi et al. (2001 p. 664) in another review of literature on privacy noted that perspectives applied to the analysis of the concepts of privacy to be:

  1. The units experiencing privacy. They go on to note “the unit experiencing privacy can be either an individual or a group, or both.”
  2. Desired – Achieved privacy. This is explained by the understanding that the concept of privacy is either seen as a subjective state or studied as an achieved state (Newell, 1998).
  3. Reactive – Proactive privacy. This is to say the control of communication and also the control of knowledge.

Furthermore, they describe the dimensions of privacy to include: physical, psychological, social and informational thus, suggesting privacy dimensions to be made up of four quadrants of the diagram as shown in figure 2.1 below..

Source: Leino-Kilpi et al. (2001)

It would be seen that in an open plan office all the dimensions of privacy as enumerated Figure 2.1 above are impinged upon; First, physical accessibility to the person is unrestricted. Secondly, the cognitive intrusions abound due to audio and visual distractions. Thirdly, it is more difficult to control social contacts for example, the choice of participants for interaction, the interaction frequency, length and content of the said interaction. Then finally, the ease with which certain private pieces of information about the person is easily accessible is a problem in open plan offices, after all, most open plan offices do not have a single route of access or a door to the work space. So, it is difficult to mark and protect ones territory and as such protect some form of private information from would be trespassers (Anjum, Paul and Ashcroft, 2004).

In the light of these perspectives, one of the definitions of privacy suggested is that “privacy is a voluntary and temporary condition of separation from the public domain” (Newell, 1998, p. 357).

Oldham, Kulick and Stepina (1991) highlighted the fact that individuals reacted negatively to environments characterized by few enclosures, closeness and high density because such environments exposed individuals to too many unwanted or uncontrolled intrusions.

It is also agreed that, the perception of the work environment leads to satisfaction or dissatisfaction with the work and the work environment. Fischer, Tarquinio and Vischer (2004, p.132 ) posit that the there are three major categories of mediating influences on workplace satisfaction and these are, “individual differences like culture, age, professional or status, organizational context and environmental features.”

All these issues could be further grouped into two; internal and external factors as relates to the individual. These two descriptions could be mapped to the two ingredients required for a need for privacy to exist i.e. a person or persons and a place. Sanders and McCormick (2002, p. 485) also point out that apart from the physical features of the built environment, “people are influenced by nonphysical features like social, cultural, technological, economic and political factors characteristic of the environment.”

External Factors

These are the place factors, usually described as the environmental or design issues which can lead to noise distractions, visual distractions, interruptions, crowding and accessibility issues (Ding, S. 2008). Due to the absence of internal walls, the low height of walls or partitions in open plan offices influences privacy; the more enclosures, the lower the people per given space and the higher the partitions, the higher the privacy perceived (Danielsson 2008; Oldham, G. R et al. 1991; Sundstrom, Herbert and Brown, 1982 etc.).

Organizational context is also considered to be an external factor. This involves the type of industry involved by the organization. For example doctors consulting rooms should provide more audio privacy compared to an architectural firms offices or design studios.

Internal Factors

This grouping is based on the person factors or what goes on within the person, the suggestion that individual differences related to but not restricted to personality traits, gender, individual experience etc. affect ones perception of, and hence the evaluation of the work environment (external or place factors). Some studies have found that variations exist across gender in perception of privacy in the open plan office (Yildirim, Akalin-Baskaya and Celebi, 2007). Also, in a different cross cultural study of privacy, Newell (1998) found that privacy was more a condition of the person thus, the duration of the experience and the change on the person as a result of the experience leads to its suspected therapeutic effect. In general perceptions and attitudes to privacy, she found that gender also played a part especially within cultures.

Maher and von Hippel (2005) and others before them showed that individual differences in the ability to handle overstimulation by the application stimulus screening and inhibitory abilities influenced the perceptions of the work environment. These inhibitory skills are cognitive in nature and such inhibitory skills are found to vary between individuals and even especially across cultures. For example, Hall (1966) points out that the Japanese are said to be content with paper walls as acoustic screens while the Dutch and Germans require thick walls and double doors to serve as acoustic screens.

Benefits of Privacy in the work Environment

Newell (1998, p. 359) relates the need for privacy to help in “maintaining healthy internal physiological and cognitive functioning subjectively described as ‘wellbeing’”. The study concluded that achieving the perceived privacy had some therapeutic effects.

On the area of performance, especially for knowledge workers like engineers, accountants, software designers, decision makers etc., auditory and visual distraction have been found to be a cause of stress and even performance impairment (Roper and Juneja, 2008). Furthermore, Oommen, Knowles and Zhao (2008) point to the likelihood of aggression and increased instances of eye, nose and throat irritations while working in open plan environments. This in turn affects productivity.

Culture

Culture is said to be the way of life of a group of people. This, among other things covers their beliefs, values, norms and rituals. Specifically, Hofstede (2009 p. 1) points out that “culture is the collective programming of the mind that distinguishes the members of one group or category of people from the others and it manifests itself in the form of symbols, heroes, rituals and values.” Earlier, an American anthropologist Edward T. Hall in his books, talked about language and especially modes of communication as a point of differentiating cultures (Hall E.T 1966; 1976). He even considered language to be the core of culture while, Geert Hofstede considers language as a part of the rituals of a particular culture (Hofstede, 2010). This goes to point out some of the existing disagreements about what culture is and even how it comes about.

Culture is thus, studied as a means of understanding or shedding light into the behavior or reactions of individuals or people groups. Edward Hall in his book; the hidden dimension writes that “people from different cultures inhabit different sensory worlds, so that experience, as it is perceived through one set of culturally patterned sensory screens is quite different from experience perceived through another.” (1966, p. 2). This highlights and explains the link between culture and perception generally and in spatial terms especially.

Classifying Cultures

There have been several descriptions and models of culture (Matsumoto and Yoo, 2006; Hall, 1966; 1970), for example, Hall (1966) alludes to contact and non-contact groups or cultures in relation to spatial meanings and preferences within people groups . This is related to the social dimension of privacy (Leino-Kilpi et al. 2001), but he especially specifies high and low context cultures according to their ways of communicating.

For the high context (HC) culture or communication for that matter, much of the information is implicit while, in the low context (LC) culture, nearly everything is explicit. He also wrote about the concept of time among cultures (Hall, 1976). Where there are polychronic (P-time) and monochronic (M-Time) cultures; the M-time society or culture would prefer to do only one thing at a time when serious i.e. for such persons, time is linear and segmented with each activity scheduled while, the individuals in a P-time culture can juggle several activities, they emphasize the involvement of people and the completion of tasks rather than schedules.

Edward T. Hall coined the term “Proxemics” which he describes as “interrelated observations and theories of man's use of space as a specialized elaboration of culture” Hall (1966 p. 1). In explaining his observations in proxemic behavior (Hall, 1963 p. 1003) he notes that “what is close to an American may be distant to an Arab.”

Many other researchers and individuals apart from Edward Hall had worked on other frameworks and dimensions of culture. Matsumoto and Yoo, (2006) lists some of these frameworks which are interestingly identified by the names of the researchers that discovered them and this list which is not exhaustive, includes;

Hofstedes (1980) with subsequent revisions and dimensions added; Schwartz (2004) who presented seven universal value orientations, Smith, Dugan and Trompenaars (1996) had two universal value orientations; House, Hanges, Javidan, Dorfman and Gupta (2003) came up with nine value orientations related to leadership; Inglehart (1997) had two attitude-belief-value orientations, Bond et al. (2004) is said to have reported two social axioms. All cited in Matsumoto and Yoo, (2006 p. 239).

The listing above does not mention each of the dimensions. The dimensions of each framework listed are found in Table 2.1 below.

Table 2.1 Six Theoretical Frameworks for Universal Dimensions of Cultural Variability

Framework

Dimensions

Hofstede’s (2001) dimensions of

work-related values

Individualism vs. collectivism

Power distance

Uncertainty avoidance

Masculinity vs. femininity

Long- vs. short-term orientation

Schwartz’s (2004) dimensions of

values

Embeddedness

Hierarchy

Intellectual autonomy

Affective autonomy

Egalitarianism

Mastery

Harmony

Smith, Dugan, and Trompenaars’s

(1996) dimensions of values

Egalitarian commitment vs. conservatism

Utilitarian involvement vs. loyal involvement

House, Hanges, Javidan, Dorfman, and

Gupta’s (2003) dimensions of

leadership values

Performance orientation

Assertiveness orientation

Future orientation

Human orientation

Institutional collectivism

Family collectivism

Gender egalitarianism

Power distance

Uncertainty avoidance

Inglehart’s (1997) dimensions of

attitudes, values, and beliefs

Traditional vs. secular-rational orientation

Survival vs. self-expression values

Bond et al.’s (2004) dimensions

of social axioms (beliefs)

Dynamic externality

Societal cynicism

Source: (Matsumoto, D and Yoo, S. H, 2006 p. 240)

National versus Organizational culture

As a society has a culture, so do organizations and such organizations employ staff who come from a particular culture(s). The organizations then require these individuals to work in offices. Apart from the culture description related to national boundaries, there is a culture that seems to characterize workplaces or organizations and this is called organizational or corporate culture.

Barney (1986) notes that like culture itself, organizational culture has many competing definitions and then goes on to suggest that a generally acceptable definition of organizational culture to be “as a complex set of values, beliefs, assumptions, and symbols that define the way in which a firm conducts its business.” (p. 657). He goes on to point the pervasive nature of organizational culture in that, it helps to define the relationship of the firm to parties it comes in contact with through its business. This simply shows that culture within the work place especially geared towards profitability or the conferment of advantages could be termed organizational or corporate culture. Generally it will be assumed that the national culture will also play a part.

Guidroz, Kotrba, and Denison (2009) from results of a study of multinational companies, claim that their study seems to point to organizational culture superseding national culture in diversity management practices. The issue in question in this thesis is not exactly a management matter but the individual perception of privacy in the open plan office environments and would suggest that both national and organizational cultures playing a part because according to (Brand, 2009) the design of the workspace or workplace communicates the corporate culture of the organization meaning, the adoption of the open plan environment can be tied to the organizations corporate culture.

Hofstedes’ Cultural Dimensions

This is a hugely popular cross-cultural model (Gerhart and Fang, 2005; Hofstede et al, 2010; Sivakumar, Nakata, 2001) currently in use, with its roots in industrial psychology (Meeuwesen, van den Brink-Muinen and Hofstede, 2009) is called the Hofstedes model of culture named after Geert Hofstede a Dutch emeritus Professor of organizational anthropology and international management in the Netherlands.

Hofstedes work highlighted the fact that culture is manifested through symbols, heroes, rituals and values. But, Hofstede argues that values form the core of culture as represented by the Hofstede culture “Onion” in Figure 2.2. Figure 2.2 shows the onion structure graphically illustrating the manifestation of culture at different levels and even the interactions therein.

As seen from Figure 2.2 above symbols, heroes and rituals are by themselves visible to all observers. It is the cultural meanings of the practices that are open to interpretation by the observer while, values are unseen or embedded within the person but they still subtly determine choices and much more (Hofstede et al., 2010). Hofstedes research studied value survey responses of similar respondents from different countries as to their approach, as related to four basic problems prevalent in most societies (Meeuwesen et. al, 2009) these included;

  1. Handling social inequalities in the society.
  2. The approach to dealing with uncertainty in general.
  3. The structure of the relationship between an individual and the group.
  4. The emotional role division between the male and females in a society.

The initial data for Hofstedes culture study came about through an analysis of International Business Machine Company (IBM) staff surveys at a time, the company was called Hermes. He utilized the responses from routine staff surveys about values and related matters to provide ratings for countries on each of what he then called the four dimensions of culture.

This was achieved by examining correlations between mean scores of questionnaire items at the level of countries. Other approaches, like analysis at the individual level did not provide much useful information (Hofstede, 2009). Later, certain studies showed the need for another dimension and this lead to the inclusion of a fifth dimension called, long term orientation.

Each dimension of culture score for a country is calculated using a scale of roughly 0 to 100 for each dimension. A dimension of culture is an aspect of culture that can be measured relative to other cultures (Hofstede, 2009 p. 6) and the higher the score of a dimension, the more that dimension is exhibited in the society or nation in question while for lower scores the opposite pole of the dimension is more pronounced. Thus, the scores are therefore bipolar (Jones, 2007)

In a 2010 book, Greet Hofstede, his son Gert Jan Hofstede and a research collaborator Micheal Minkov reviewed earlier works, alongside their recent studies and added a sixth dimension called indulgence versus restraint (IVR) to the previously known Hofstedes five dimensions of culture. The sixth dimension was largely as a result of the work of Micheal Minkov (Hofstede et al., 2010).

The six dimensions of Hofstedes cultural model now include power distance (PDI), individualism (IDV), masculinity (MAS), uncertainty avoidance index (UAI), long term orientation (LTO), and the recently added indulgence (IVR).

Power distance (PDI).

This indicates the degree of inequality that exists and is accepted among the persons with and without power i.e. the leadership versus the followership respectively as normal and legitimate in any given society. If the power distance scores are high, it indicates a pyramidal or hierarchical system where the power is resident at the top while, lower scores indicate greater equality suggesting power is shared and spread within the group.

Individualism (IDV).

This is related to the sense of community that exists within the group. High scores indicate a lack of, or even weak connections the beyond self, family and friends while, lower scores indicate collectivism or group cohesion and such societies take more responsibility for each others well being. Yintsuo (2007) notes that in-group members and in-group goals have priority over ones personal interests. There is emphasis on team work and inclusion in collectivist societies (Guidroz et al., 2009). The Chinese are known to be highly collectivist (Johnston, Warkentin and Luo, 2009).

The individualism versus collectivism dimension is argued to be the most studied of all dimensions of culture and linked to many psychological differences across cultures (Matsumoto and Yoo, 2006).

Masculinity (MAS).

This sometimes represents the position of the society on traditional male and female roles. In actual fact, masculinity scores indicate in some way levels to which in a society the male and female emotional roles are distinct. A strongly masculine culture group would be more assertive, tough and emphasize material success and the women would be expected to be modest, tender and concerned with the quality of life (Hofstede et al., 2010). Lower scores indicate the blurring or overlap of emotional gender roles with males and females all being modest, tender and concerned with the quality of life. It would be seen that both sexes work side by side in all spheres of endeavor in such societies.

Uncertainty avoidance index (UAI).

This is said to be the feeling of anxiety members of the society feel in uncertain or unknown situations. High scores signify avoidance of ambiguous situations when possible. Low scores show a group that enjoys novel events and values differences, usually having few rules in order to allow people discover the truth for themselves. Uncertainty accepting cultures are more tolerant of opinions different from what they are used to.

Long term orientation (LTO).

This dimension was added later to make up the five dimensions of culture. It was earlier called the Confucian dynamism. It is a reflection of value attached to tradition. High scores indicate a high long term value while lower scores the exact opposite. Societies on the low side are said to be short term oriented. It was observed that the nations in Asia with an influence of Confucian ideals rank highly on this dimension. In fact, Jones (2007) considers this dimension to be an attempt to fit uncertainty avoidance into Asian culture. The discovery and addition of this dimension was as a result of a study to examine western bias in the initial Hofstede survey (Yeh and Lawrence, 1995).

Indulgence (IVR).

The most recent addition to the Hofstede paradigm reflects the societal gratification of basic and human natural human desires related to enjoying life and having fun and its complement, restraint, reflects a conviction that such gratification needs to be curbed and regulated by strict social norms (Hofstede et al., 2010). It is related to subjective wellbeing and interestingly, the countries that score the highest are some of the poorest countries in the world.

Criticism of Hofstedes Model of National Cultures

In a paper titled “Hofstede – Culturally Questionable?” Jones, (2007) points out that Hofstedes work on cultural dimensions is controversial with a large number of protagonist and antagonist. Some of the issues raised include; relevance, cultural homogeneity, political influences, one company approach, the dimensions being out-dated, some suggest there are too few dimensions and then for others questions of statistical integrity. This list shows the whole gamut of argument used by the antagonists.

In the following paragraphs and under the discussion on the adoption of the Hofstede paradigm, some of these issues will be discussed.

According to some researchers (McSweeney, 2002; Myers and Tan, 2002) there is no necessary alignment between culture and the nation-state. They propose that, researchers should adopt a more dynamic view of culture; one that sees culture as contested, temporal and emergent.

In cautioning on the use of the Hofstede culture model in accounting research, Baskerville, (2003 p. 7) pointed out that “incorporating and equalizing cultures with nations minimized the variety in the units of analysis.” She again went on to say that due to Hofstede been influenced towards psychology; he really studied national character, rather than national culture. These views are just a few of the criticisms leveled against the model of culture proposed by Hofstede.

Secondly, a look at the Hofstedes nation list shows that some countries are grouped together, for example, Arab countries, East Africa and West Africa and in some cases along racial lines; like in providing scores for South African Whites (Hofstede et al., 2010).

Again, West Africa, which is essentially a grouping of 16 countries covering an area of almost 5 million Square kilometers with 12 countries using French as their official language; showing French colonial influences. The rest employ English also an indication of English colonial influences. It would portend a great generalization to group such nations under a single cultural value score.

Nigeria, one of the 16 countries has 250 ethnic groups and a catalogue of 521 indigenous languages. This language and cultural diversity, suggest a gross generalization of the cultures of Nigeria, not to talk of the whole of the West African region if all 16 nations are grouped together.

In recent publications (Hofstede et al., 2010), however, Hofstede and his collaborators are beginning to provide more nation specific scores instead of regional scores.

In response to several criticisms leveled at his work, to which he has provided replies. Greet Hofstede asserts that several studies indicate that the country differences described by the dimensions are basic and long term (Hofstede, 2009; Hofstede et al, 2010).

Adoption of Hofstedes paradigm

Despite the problems observed, Hofstedes model of culture has proven to be the model of choice to a lot of researchers across a lot of fields. In fact it is said that Hofstede is one of the most cited authors (Jones, 2007; Sivakumar and Nakata, 2001). Again Jones (2007 p. 6) points out that the Hofstede paradigm “has many appealing attributes” and he also notes that the points of agreement among the bulk of the Hofstede paradigm protagonists include the frameworks relevance, rigor and relative accuracy.

Cross-cultural studies in several areas are said to have validated the framework (Hofstede et al., 2010) along with other papers in information technology, business and the social sciences (Johnston, Warkentin and Luo, 2009).

More of these results will be discussed under studies employing the Hofstede paradigm.

Culture’s Effect on Privacy Perception

Individual differences exist in the perception of privacy in open plan office environments; this has been observed from different studies. Ding (2008 p. 402) claims that privacy is a “psychological perception, other than describable and quantifiable phenomenon”. Some studies have found some differences in response (or perception) to open plan offices across gender; with the males responding more positively to open plan offices this, was reported by Yildirim, Akalin-Basaya and Celkebi (2007).

Other researchers, for example, Fischer et al. (2004) showed that, the individual’s sense of identity or self schema affected his or her perception of the work environment.

The development of the schema is formed from feedback; positive or negative gotten from the community. This community is made up of supervisors, colleagues in the work place and then, the judgment made by the individual of the work environment. This study highlighted the place of the community on the perception of the work environment.

Similarly, Markus and Kitayama (1991) showed that people in different cultures hold different construal’s of self, of others, and of the interdependence of the two (self-others) thereby influencing and determining individual experience, including cognition, emotion, and motivation. It is therefore not surprising that, Hall (1976 p. 13) considered culture to be “a series of situational models of behavior and thought”. He pointed out also, the pervasive nature of culture, as it influences practically every thing by imposing a kind of blindness to certain issues.

Again, some researchers like Nisbett and Miyamoto, (2005) whose work on perception, stimulated by earlier research on the area of human cognition, showed that inferential processes are affected by culture. They suggested that perception can no longer be regarded as consisting of processes that are universal but would vary as influenced by culture. Their work points to the subtle import of cultural differences in the way stimuli are compared. Furthermore, Wu and Keyser (2007) pointed to the observation that the different cultural patterns of independence and interdependence among westerners and easterners respectively, affect how they remember and perceive events.

Researchers like (Newell, 1988; Hall, 1966) pointed out that in certain cultures; the concept of privacy is almost none existent when compared to the understanding and use of the term in other cultures. As evidenced by the absence of words to that effect in the vocabulary of such groups.

Kaya and Weber (2003) studied cross-cultural differences in the perception of crowding and privacy regulation among Turkish and American students in similar university residential surroundings. Their findings showed that, there were differences in privacy desires between cultures and gender.

So, it is clear that there is to be expected some variations in privacy perception of users of open plan offices with respect to culture.

Some Studies employing Hofstedes paradigm

Johnston et al. (2009) in preparation of their work on information security showed that the individualism versus collectivism dimension of culture plays a very important part in forming privacy expectations among staff as regards security policies of their companies. Their study was designed to cover working professionals in the insurance and other industries in the United States of America and China. They expected that the expectation created by culture would affect privacy perception.

Deschepper et al., (2008) conducted a study examining the import of cultural dimensions on antibiotic use in Europe. They used culture dimensions scores for the 19 European countries already put up by Hofstede and then compiled the data on antibiotics use taking note of prescription use and self medication in three different studies carried out around 1997 - 2003. The data obtained was analyzed using a simple correlation analysis.

The results of the study showed significant positive correlation between power distance indexes (PDI) with the prescribed use of antibiotics in all the three different studies. It was also correlated with self medication in one of the studies. Also, a significant positive correlation was found with uncertainty avoidance (UAI) in two of the studies. Interestingly, masculinity (MAS) was not significantly correlated except for when GDP was controlled and only in one study was this found to be true.

Nikoomaram, Kavousy, Kangarluei and Bayazidi (2010) used Hofstedes model to investigate cultures impact on earnings of companies registered on the Tehran stock exchange. They found the individualism dimension to have the most impact on earnings management but the whole of the dimensions only accounted for 12% of the variation in earnings of the companies studied. They administered the Hofstede values survey module (VSM) questionnaire to company staff and then used responses to calculate Hofstede cultural values which were then used in the analysis. In this study, they employed the use of least squares regression to test the predictive capability of the four dimensions of Hofstedes culture dimensions.

Meeuwesen et al. (2009) investigated how cross-national differences in medical communication can be understood from the first four of Hofstedes cultural dimensions namely; power distance (PDI), individualism (IDV), masculinity (MAS) and uncertainty avoidance index (UAI) alongside national wealth i.e. the gross national product of the nations studied. In the study they sampled doctors and patients. Here also, a stepwise regression was used to study the predictive capacity of Hofstede cultural values to predict medical communication styles.

As related to the office environment, there is near absence of papers that examine the import of culture or specifically cultural values in perceptions or attitudes towards open plan offices.

Challenges of Cross-Cultural Research

The general difficulty in cross cultural research has been documented in literature (Jones 2007; Sekaran 1983 etc.) In particular Jones (2007) outlines several salient issues faced by the cross cultural researcher namely; problems of definition especially of the word “culture”. Next, is what he calls a methodological simplicity citing the ethnocentric pattern focus which, leads to bias and misinterpretation for example, equating cultures and nations. Another key issue related to methodological simplicity is the lack of a multidisciplinary approach.

Equivalence is also listed as a challenge by Jones (2007 p. 3) with four aspects namely: functional, conceptual, instrument and measurement equivalence. Functional equivalence should stem from sameness in roles of the sample population for examples, student be compared to students, conceptual equivalence relates to “cultural utility of behavioral or attitudinal constructs” meaning the attributes should exist in the cultures studied. The instrument and measurement equivalence which refers to the degree to which measures used to collect data in different cultures are equally valid and reliable through the elimination of non cultural differences (Matsumoto and Yoo, 2006). For example, the idea that a research questionnaire may be delivered in multiple languages means exact meanings could be lost in translation. Thus, back translations are said to help in this regard.

Matsumoto and Yoo, (2006) record another side of the challenge to be the reluctance in certain cultures to selecting extremes, this would relate to scales used in attitudinal research questionnaires. This is usually called the response bias which can be of socially desirable responding which has two variations namely; extreme response bias and then, self deceptive enhancement and impression management. Another is acquiescence bias also called hospitality or courtesy bias (Sekaran, 1983). Sekaran (1983) also records the sucker bias where in some cultures any foreigner is considered fair game and that idea is carried over to responses presented in questionnaires or interviews. All these, highlight the challenges of cross cultural research, the importance of pilot studies and especially, the need for local collaboration.

Sampling issues in Cross cultural research

There are said to be three (3) types of cross cultural research and these are based on the selection and treatment of the respective samples i.e. nations or cultures. In the Hand book of cross- cultural psychology John, Ype and Janak (1997, p. 262-263) note that sampling of cultures have been found to fall into 3 categories.

The first and fairly common approach is the convenience sampling which as the name suggests stems from the researcher being from such a culture and or familiarity with some collaborator(s) in the other culture(s). Sekaran (1983) describes such a scenario as opportunistic sampling.

The second approach is the systematic method because the cultures “represent different values on a theoretical continuum” In this regard, Sivakumar and Nakata (2001) proposed a method to aid in more systematic multi-country sample selection using the Hofstede framework by developing an algorithm that calculated indexes and then ranked with the top most combination, which are the countries furthest apart on each dimension as being the most ideal combinations to strengthen hypothesis testing.

Thirdly, there is the random sampling wherein a large number or cultures are randomly sampled this is sometimes called a pan cultural approach. The pan cultural approach can be both time and resource consuming as such relatively few of such studies have been carried out.

Finally, there is the issue of subject sampling where the subjects sampled should be similar in terms of relevant background characteristics as mentioned earlier, students in one culture should be compared to students in another culture and not compared to teachers in order to result in valid cross cultural comparisons.

Research Question

In the course of the study, it is expected that the variation in privacy perception among open plan office users in the three countries covered by the study would be investigated and secondly, an analysis will be conducted to find out if there exists a relationship between privacy perception in open plan offices, and culture or cultural values as represented by the Hofstedes cultural dimensions.

Research Hypothesis

According to Zalesny and Farace (1987) the physical setting of work provides information of a person’s social position. This could be a potential source of dissatisfaction with open plan offices by individuals on the staff in higher echelons of companies in cultures with a high power distance (PDI).

This study will attempt to find out how the privacy perceptions will vary across the national cultures of China, Nigeria and the United States of America, i.e. if it varies at all and then examine the relationship of privacy perception to culture using Hofstedes culture dimensions.

Main Hypothesis

The first hypothesis states that there a variation in privacy perception of open plan offices across office workers in China, Nigeria and the United States of America.

In the light of the foregoing, the main hypothesis of would be that, there is a significant relationship between cultural values and privacy perception in open plan offices.

Sub-hypotheses

  1. There is a significant relationship between power distance and privacy perception in open plan offices.
  2. There is a significant relationship between uncertainty avoidance and privacy perception in open plan offices.
  3. There is a significant relationship between individualism and privacy perception in open plan offices.
  4. There is a significant relationship between masculinity and privacy perception in open plan offices.
  5. There is a significant relationship between long term orientation and privacy perception in open plan offices.
  6. There is a significant relationship between indulgence and privacy perception in open plan offices.

Research Motivation

The study is expected to examine the variations in privacy perception across national cultures of China, Nigeria and the United states.

The loss of privacy even though thought to be relative, could be said to lead to dissatisfaction with the work place and if a staff, is dissatisfied with their Jobs this in turn affects the organizations general performance (Veitch et al. 2007; Vischer 2006).

Secondly, the import of the 6 Hofstedes cultural dimensions especially the Power distance index (PDI), Individualism (IDV), Masculinity (MAS), Uncertainty avoidance Index (UAI), long term orientation (LTO) and the newly added Indulgence versus Restraint (IVR) scores of each country could then be compared to see if they will provide an indicative and predictive capacity to determine privacy perception in open plan offices.

Thirdly, the absence of any studies attempting to study the relationship between culture and satisfaction or differences in perceptions in the open plan offices despite a number of empirical studies that suggest that culture could play a part in acceptance or the maximization of the benefits of the open plan office environment. Newell (1998) noted that there was a fit of cross-cultural desire of privacy to the universal value scheme proposed Schwartz and Bilsky (1987) namely; self-direction, universalism, benevolence, tradition, conformity, security, power, achievement, hedonism and stimulation.

Statistical Analysis Tool

The analysis tool to be employed in analyzing the resulting data is the Statistical Package for Applied Sciences (SPSS). This commercial software was first issued as Predictive analytics SoftWare (PASW) now a proprietary product of IBM Company after an acquisition of the original vendor company.

SPSS is not just capable of statistical analysis but also data management and data documentation and has compatible version for operation systems like Windows, Linux/UNIX and Macintosh. It has a relatively easy to use graphical user interface (GUI) and can also accept syntax codes. SPSS can read and write data from spreadsheets and databases. SPSS outputs can be exported and written to external applications like the popular Microsoft office. SPSS now boasts of its 19th version stable version.

The IBM SPSS is also very common statistical analysis tool (Field, 2006). There are other free open source applications and commercial software also available and used in cross cultural studies.

CHAPTER 3:METHODOLOGY

Design of the Study

According to Sundstrom et al., (1982) research in open plan offices has been conducted using attitude surveys employing questionnaires. Sometimes, the data is elicited as pre and post occupancy surveys and at other times retrospective surveys. In some cases, however, like Ding (2008); a structured interview of users was the approach used.

This study will examine the privacy perception of open plan office users in the three different countries through the use of a questionnaire. This is a simple, time saving and cost effective strategy to acquire data about privacy perception from open plan office users across the nations (cultures) proposed for the study.

The research aims were spelt out in the research hypothesis and motivation in the previous chapter. Also covered in this chapter is the identification of the 3 nation samples; China, Nigeria and the United States of America.

Research Setting

The plan was to survey atleast 50 respondents from each country working in an open plan office. (Field, 2005) recommends 10 to 15 samples per predictor (criterion) for a regression analysis. The questionnaire was designed to be simple, preferably, administered in a native or official language of the country of choice. This led to the decision to use English for Nigeria and USA while Mandarin was used for China.

Samples

The choice of samples would be described as opportunistic and based on perceived convenience in that, being a Nigerian student in China provided the relative ease with which relevant cross cultural data could be collected. This informed the choice of two of the countries to be studied namely, China and Nigeria. The American sample was a later addition to an otherwise two nation study just to offset the similarities between Chinese and Nigerian cultural values based on the Hofstede cultural framework.

The people’s republic of China

The people’s republic of China has a population of over 1.3 Billion people and is said to be the most populous nation of the world. It recently became the second largest economy of the world displacing Japan while trailing the United States of America. China covers an area of 9.6million km2 and a population density of 139.5 persons per km2 and has Gross Domestic Product (GDP) per capita of $6,567 with literacy levels at 93.3% (Wikipedia, 2010).

China also has the enviable record of being one of the longest continuous cultures that has resided within roughly the same geographical area. Its cultural dimension scores are shown in Figure. 3.1

The Federal Republic of Nigeria.*

The federal republic of Nigeria is a West African country and has a population of over 150 Million persons. It is said to be the largest homogeneously black nation of the world. The country covers a total area of 923,768km2 and has a population density of 167.5 persons per km2. Nigeria’s Gross Domestic Product (GDP) per capita is put at $2,249 with literacy levels at 72% (Wikipedia, 2010).

Nigeria has a large religious and ethnic mix; it is almost evenly split along religious lines with the two majority religious groups being Christians and Moslems alongside other religious systems. It has 500 plus languages, spread across 251 ethnic groups. Its Hofstedes cultural dimension scores are shown in Figure. 3.2

*All scores are for West African except for LTO and IVR. Nigeria is one of the 16 countries in the geographical region called West Africa

The United States of America.

Has a population of over 309 Million. It is said to be the world’s largest economy. It covers an area of 9.6million km2 and has a population density of 32.2 persons per km2 and a GDP per capita of $46,381with literacy levels at 99% (Wikipedia, 2010). Its cultural dimension scores are shown in Fig. 3.3

Comparative Views on the Three Nations Cultural dimensions

Figure 3.4 below shows the Hofstedes culture scores for the three nations chosen for this study, It could be seen that that both Nigeria and China score highly on the power distance dimension (PDI) while the United States of America scores high in individualism, while the other are more Collectivist. The masculinity score for Nigeria was the lowest but all around the average mark. The Uncertainty avoidance index scores show Nigeria to be slightly higher than the others. China scores high in the long term orientation dimension (LTO) which is related to Confucian ideals and Nigeria scores highly in the Indulgence dimension (IVR) score closely followed by the United States of America.

Questionnaire Design

The instrument of choice was a questionnaire which was designed to elicit the privacy perception of office workers in the three nations covered by the study.

The research instrument was designed to be a scenario based questionnaire i.e. it required all respondents to imagine they were required to work in an open plan office. The questions in the second part of the questionnaire were adapted from a set of factors related to satisfaction with privacy and acoustics in a Cost–effective Open–Plan Environments (COPE) project study by Veitch et al. (2007).

In that study, they measured environmental satisfaction using a questionnaire, measurements of the respective work stations and developed a model that provided a link between environmental satisfaction and job satisfaction using structural equation modeling (SEM). A factor analysis of the data resulted in a three factor solution namely; satisfaction with privacy and acoustics, satisfaction with ventilation and temperature and satisfaction with lighting. The variables of privacy and acoustics were 10 in all and they had the highest factor loading scores. These items where then taken and developed into a scenario type questionnaire with nine items for privacy perception. See Appendix A.

The initial plan was not just to elicit the privacy perception but to also obtain fresh scores for the cultural dimensions scores for which Hofstede says there must be a minimum of 50 participants for the survey results to be reliable (Hofstede et al, 2010). This idea was discarded after the pilot study as, the questionnaire would have included over 50 questions and all pilot participants complained about the length of the questionnaire. In fact, the first set of pilot questionnaires had 3 sections and 56 questions in all; the general information part, the privacy perception part and then the cultural dimension part which was the values survey model 2008 (VSM 08) developed by Hofstede and made available in several languages for free to academic researchers on Greet Hofstedes website - http://www.geerthofstede.nl/research--vsm/vsm-08.aspx.

Complaints about the length of the questionnaire led to the abandonment of the third part of the questionnaire which was the VSM 08. This decision was supported by that fact that the culture value scores were readily available for researchers from several sources and several studies for example Deschepper et al. (2008) have simply enjoyed the benefit of the already widely available prepared cultural values data instead of recalculating the cultural values all over.

Scales

A Likert scale was used to measure privacy perception since the Likert scale is appropriate for measuring levels of attitude or impression and can provide an effective measure of user satisfaction (Gliem and Gliem, 2003; Heim 2007). A scale of 1 to 7 was used to elicit perceptions of privacy in open plan offices. This is simply because it has been found in some studies to have a higher sensitivity than 4 or 5 point scales (Sekaran, 1983). The seven point scale is also very common in attitudinal studies for example (Anjum et al., 2004; Veitch et al., 2007).

Equivalence

To ensure reliability and validity of the translated questionnaire, back translation and bilingual techniques are recommended to be used during the translation of the questionnaire into Chinese and subsequent testing of the questionnaire (Johnston et. al, 2009; Newell, 1998).

The questionnaire was developed in English and pretested before its translation to Chinese. The translation was done by a graduate student in the Industrial engineering department who was proficient in English and it was also pretested before the large scale deployment. For the Chinese translation of the questionnaire, see appendix B.

Measurements and Analysis

Taking measurements of the work spaces of the respondents would have provided more information to enable a favorable comparison of the conditions of the work spaces of respondents across cultures. Measures like the ambient noise levels, heights of partitions, space allocation or density etc. would have been taken.

Nationality, gender, age range, rank, tenure and prolonged exposure to other cultures will be elicited and statistically analyzed if necessary. The main area of concern being the perception of privacy in open plan offices as elicited by the scenario based questionnaire used in this study.

The study had to be carried out as a scenario based questionnaire study since it would have taken a longer time and a large amount of funds to complete a more direct study.

Privacy perception Part

The second section of the questionnaire was designed to examine the perceived satisfaction with an open plan office environment and the 9 item questionnaire had the following questions.

Q1. How would noise from other people’s conversations affect you while you are at your workstation? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q2. How would frequent distractions from colleague’s movements affect your work? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q3. How would frequent distractions from mobile phone ring tones, copier, fax etc. noises affect your work? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q4. What degree of importance do you attach to enclosure of your work area by walls, screens or furniture? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q5. What level of visual privacy do you think an open plan office with chest high walls/partitions provide? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q6. How far apart, do you recommend (roughly in meters), the distance between you and others, if you are to work in an open plan office? The options were a list of distances in meters from 1 meter to 7 meters in 1 meter increments.

Q7. How would you rate the level of audio privacy (for confidential speech) required for your day to day conversations in your workspace? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q8. How do you rate the acceptable level of background noise (e.g. air conditioner hum etc.) you hear at your workstation? The options were a Likert scale from 1 indicating a low, 4 average and 7 high.

Q9. How important is the freedom to drastically alter aesthetic appearance of your workspace? The options were a Likert scale from 1 indicating not important, 4 average and 7 very important.

Coding

The responses to the 9 item privacy perception part were coded automatically if the respondent completed an online questionnaire, for the hard copy questionnaire the results had to be manually entered into spread sheets. The responses to questions 8 and 5 were reverse coded to reflect affect of the privacy issues in an open plan environment where it is assumed that audio and visual distractions would abound. The level of enclosuee will be zero as the office will be assumed to be without interior partitions of any kind.

Statistical Analysis

The general statistical distribution of the respondents across the three countries where respondents elicited would be analyzed and obtained results presented with tables and graphs here necessary.

In line with research practices, an exploratory data analysis will be done to determine the nature of the data given that the parametric or non parametric nature of the data determines the best kind of test to produce reliable results (Fields, 2005; Heppner and Heppner, 2009; Meeuwesen et al., 2009).

Analyzing cross cultural data involves the use of alot of techniques both quantitative and qualitative. Several researchers have noted the preponderance of multivariate techniques for the study of cross cultural data (Sekaran, 1983). The general complexity of cross cultural research would naturally require multivariate approaches to the data analysis as, the variables are usually large in number. In the privacy perception part, 9 questionnaire items are required to analyze the privacy perception in open plan offices. Thus, only the Multivariate analysis of variance (MANOVA) has the capability to simultaneously compare the means of the responses to the questionnaire.

The analysis would mainly involve a multivariate analysis of variance (MANOVA) between cultures or nations, gender and other values on perceived privacy items of the questionnaire and then a regression analysis to find out the predictive capability of Hofstedes cultural dimensions scores, on perception of privacy in open plan offices. The outcome or dependent variable will be the perceived privacy while the predictors or independent variables will be the Hofstedes cultural dimensions of power distance (PDI), individualism (IDV), masculinity (MAS), uncertainty avoidance index (UAI), long term orientation (LTO), and indulgence (IVR).

CHAPTER 4

ANALYSIS AND RESULTS

This chapter presents and discusses the results obtained from the designed and administered questionnaire. In line with the methodology earlier proposed, a preliminary data analysis preceded the subsequent treatment of the data to enquire if the data lends itself to the desired analysis. Some of the important tables obtained are presented with graphical presentation of results. The complete results are attached in the appendices.

Survey Procedure

The initially designed questionnaire was first administered as a pilot test among 4 persons 2 Americans and 2 Nigerians and they were required to complete the questionnaire and make comments that would help improve the questionnaire. Two main complaints where received; one of the American respondent noted that the question (Q6) on the preferred working distances between work stations in an open plan office, it was difficult for Americans to relate to measurement of distances in meters but in feet. The second was the length of the questionnaire. Initially the Hofstede cultural values were to be recalculated but the VSM 08 module had 30+ questions this would have made the questionnaire have 50+ questions. After the translation, the Chinese questionnaire was tested by 6 individuals and no problems where found.

The questionnaire was then administered in a several formats mainly hard copied versions and then online versions using online free survey portals. The first was the Google documents survey facility but with the Google restrictions on the Chinese mainland, respondents could not access the questionnaire on the Chinese mainland. The whole questionnaire was redone with an English and Chinese version on another free online survey platform called kwiksurvey – (www.kwiksurveys.com)

The data was randomly and anonymously collected and confidentiality was guaranteed and a statement to that effect was clearly written in the introduction to all the questionnaires online or hard copy. At the close of response collation, 247 persons had valid responses with the 10 others rejected for gross incompleteness or non discrimination in completing questionnaires for example, some respondents just selected or marked the maximum score throughout. The distribution of the respondents by nationality is 104 Chinese, 75 Americans and 68 Nigerians. Some of the data collected in the general information section is analyzed below under the general statistics.

General Statistics

The type of work involved by the respondents were classified into 6 groups namely; Administration, secretarial, technical, professional, management and others Table 4.1 shows the distribution across the three nations with the largest group being in the Technical cadre followed by the professional cadre. The grouping listed as others include teachers, journalists, businessmen etc.

Table 4.1 The Job distribution of the respondents

Job

Administration

Secretarial

Technical

Professional

Management

Others

China

27

3

23

14

20

17

USA

11

3

9

19

9

24

Nigeria

6

2

29

17

6

8

Total

44

8

61

50

35

49

The age distribution of the respondent across the three countries surveyed is presented in Table 4.2 with the mode being the respondents in the 25-29 age range as shown in Figure 4.1. Most respondents in this category had worked for less than 5 years.

Table 4.2 Age distribution of respondents

Age Range

<20

20-24

25-29

30-34

35-39

40-49

50-59

60+

China

1

20

45

25

6

6

1

0

USA

1

15

25

14

7

5

6

2

Nigeria

0

2

13

24

20

7

1

1

Total

2

37

83

63

33

18

8

3

The distribution of males and females surveyed is presented in Figure 4.3 below. In each of the countries surveyed, the number of females to males was almost an even split except for Nigeria where the female respondents were just about 15% of the total respondents. Overall, there were 58% males and 42% females surveyed across the three nations.

Table 4.3 Gender distribution of the respondents

China

USA

Nigeria

Totals

Males

45

40

58

143

Females

59

35

10

104

104

75

68

247

The description of the respondents’ current work place as in Table 4.4 shows that of the 247 persons sampled across the three nations, 120 individuals described their workspaces as open plan offices which mean there is an absence of floor to ceiling walls and personal enclosures assigned to that individual. This corresponds to 49% of the total sample. 57 persons surveyed indicated their work spaces were cubicles or specifically had low walls or some form of partitions this corresponds to 23% of the entire sample. Then 7% of the respondents described their offices as places with thin or see through walls. This implies corresponds to 17 individuals and those in the traditional or cellular with floor to ceiling partitions dimensioned for an individual were just 50 corresponding to 20% and interestingly, 3 respondents who were all Americans worked from home of simple did not have a workspace at work equivalent to 1% of the total sample.

Using the general descriptions of offices namely; open plan offices and traditional or cellular offices implies that, 72% of all respondents worked in some form of open plan office while 27% occupied cellular offices and 1% worked from home.

Table 4.4: Workspace description across the three countries

Open Plan

Low wall/Partitions

Thin or See through Walls

Fully enclosed with doors

Home office

China

46

38

11

9

0

USA

37

17

5

13

3

Nigeria

37

2

1

28

0

Total

120

57

17

50

3

Interestingly, the choice of an ideal office showed that most individuals would prefer to work in a fully enclosed office or some variation of the same. This corresponds to 62% of the entire respondents across all the nations while 38% still chose some form of open plan office i.e. cubicles or landscape offices. It is not surprising as Oldham et al. (1991) found that for certain routine tasks are accomplished faster in company of others while certain complex task require isolation, suggesting task complexity may affect the choice of an ideal work place.

Table 4.5: Choice of an “ideal office” by the respondents

Open Plan

Low walls/Partitions

Thin or see through walls

Fully enclosed

China

22

21

16

45

USA

19

19

9

28

Nigeria

2

12

6

48

Total

43

52

31

121

A break down of the choice by nation in Table 4.5 shows that the choice of some form of open plan office was 41% for Chinese, 51% for Americans and 21% for Nigerians while the choice of a fully enclosed work space or traditional offices had 59%, 49% and 79% for Chinese, Americans and Nigerians respectively this showing that Nigerians had the highest preference of cellular offices.

The reasons for the choice of office were also elicited in the survey and results for the landscaped office which is the typical open plan office environment across all the three nations and all other types of offices or work spaces are presented below.

Table 4.6: Reasons of choosing open plan offices

China

USA

Nigeria

Total

Personal Privacy

1

0

0

1

Client Privacy

0

0

0

0

Minimal Distraction

0

1

0

1

Interaction with colleagues

14

17

2

33

Professional Status

1

0

0

1

Others

6

1

0

7

22

19

2

43

Table 4.6 shows the main reason for the choice of an open plan office was the opportunity to interact with colleagues as shown in the Figure 4.2, of all that chose the open plan office, 77% cited the opportunity for interaction as a reason for their choice. Other reasons given included the cost savings on the part of an entrepreneur or business and for some, the casual atmosphere, for yet another it is the likelihood of having access a window etc.

Figure 4.2 Pie chart showing reasons for choosing open plan offices

The reasons for the choice of low walled or partitioned workspaces or cubicles were also the opportunity for interaction with colleagues and then the opportunity to minimize distractions as shown in Table 4.7.

Table 4.7 Reasons for choosing low walled partitions/cubicle

China

USA

Nigeria

Total

Personal Privacy

3

2

1

6

Client Privacy

0

1

0

1

Minimal Distraction

4

3

5

12

Interaction with colleagues

11

12

5

28

Professional Status

3

0

1

4

Others

0

1

0

1

21

19

12

52

These results indicate that will the walls or partitions some measure of privacy was gained and distractions minimized. But the office design still enhanced interaction. The breakdown is show graphically in Figure 4.3.

Figure 4.3 Pie chart showing reasons for choosing low walled or partitioned work spaces

The choice of the thin or see through walled office was the least selected, although it is a fully enclosed office, it has the weakness of low visual or audio privacy due to the transparent nature of the wall material and the poor acoustic insulation of the walls in some cases. Of all the respondents’ across the three nations, just 31 persons described it as an ideal office. The result is presented in Table 4.8 below.

Table 4.8: Reason for choosing work spaces with thin or see through walls

China

USA

Nigeria

Total

Personal Privacy

3

5

0

8

Client Privacy

0

0

1

1

Minimal Distraction

10

3

2

15

Interaction with colleagues

1

1

2

4

Professional Status

1

0

1

2

Others

1

0

0

1

16

9

6

31

The reasons of the choice of the thin or see through walled office were primarily minimized distraction, then personal privacy issues and some still though of interaction with colleagues which seems to be strange except it is a shared office space. The breakdown is graphically presented in Figure 4.4.

Figure 4.4 Pie chart showing reason for choosing work spaces with thin or see through walls.

Almost half of the total sample surveyed chose the fully enclosed, cellular or traditional type of office the reasons for this choice was for minimized distraction, personal privacy and also their professional status and probably related to task complex.

Table 4.9 Reasons for choosing fully enclosed/cellular offices

China

USA

Nigeria

Total

Personal Privacy

17

9

12

38

Client Privacy

0

1

3

4

Minimal Distraction

19

13

28

60

Interaction with colleagues

2

1

0

3

Professional Status

7

4

5

16

Others

0

0

0

0

45

28

48

121

Table 4.9 above and Figure show that of all that chose the fully enclosed office as the ideal workspace, 49.5% cited minimal distraction as the main reason for choice, 31% chose personal privacy as the reason for their choice while, 13% considered it befitting of there professional status. The relation to daily task complexity was not studied.

Figure 4.5 Pie chart showing reasons for choosing cellular or traditional offices

Exploratory Data Analysis

The first objective of the exploratory data analysis was to examine the collected data to check for errors that might have arisen during manual coding and even inconsistencies in the data collected. This is possible because missing values, minimum and maximum scores are reported and as such outliers are easily spotted and eliminated or corrected if necessary.

Secondly, the statistical analysis methods chosen usually depend on the characteristics of the data because commonly used statistical tests are grouped in to two namely; parametric and non parametric tests. Parametric tests are usually applied to normally distributed data and non parametric tests to non-normally distributed data or categorical data. See appendix C for frequency plots of the research data.

Using SPSS, it is relatively easy to examine the normality of the dataset collected in the case of this study there are two groups of data. The first is the Hofstede cultural values scores shown in Table 4.10 below. The second is the privacy perception section of the survey instrument which is found in Table 4. 11. There are several tests for the normality of data available and several difficulties in interpreting the meanings, the most common test for data samples less than 2000 is the Sapiro-Wilk test which tests for the normality of a sample as a null hypothesis and expresses the answer as a measure of significance. Usually p > 0.05 rejects the hypothesis and the data is assumed to be normally distributed however, there are situations where such answers may not be found but the data could be assumed to be normal. The results of the Kolmogorov-Smirnov and Shapiro-Wilk tests are found in appendix D.

Table 4.10 Descriptive data for the 3 Nation Hofstede cultural values

PDI

IDV

MAS

UAI

LTO

IVR

N

Valid

247

247

247

247

247

247

Missing

0

0

0

0

0

0

Mean

67.03

41.56

59.28

41.47

48.11

53.88

Std. Error of Mean

1.141

2.082

0.533

0.653

2.138

1.67

Median

77

20

62

46

26

68

Mode

80

20

66

30

87

24

Std. Deviation

17.926

32.714

8.372

10.26

33.603

26.249

Variance

321.345

1070.22

70.096

105.274

1129.19

689.01

Skewness

-0.844

0.859

-0.881

-0.06

0.256

-0.16

Std. Error of Skewness

0.155

0.155

0.155

0.155

0.155

0.155

Kurtosis

-1.276

-1.272

-1.061

-1.719

-1.864

-1.792

Std. Error of Kurtosis

0.309

0.309

0.309

0.309

0.309

0.309

Range

40

71

20

24

74

60

Minimum

40

20

46

30

13

24

Maximum

80

91

66

54

87

84

Another approach to the normality question is to obtain the frequency plots of the items of the questionnaire and the typical bell shaped curve provides a good measure. Two characteristics of the normal curve are the kurtosis and the skewness which represents the characteristic of the “peak” and “spread “of the normal curve and according to Heppner and Heppner (2009, p. 158) “the closer to zero the better, and less than the absolute value of 2 is desirable” for the kurtosis and skewness to be tolerated as normal thus, a look at the tables of both groups of data show that the data could be assumed to be normally distributed.

Table 4.11 Descriptive statistics for the privacy perception section

Q1

Q2

Q3

Q4

Q5_R

Q6

Q7

Q8_R

Q9

N

Valid

247

247

246

246

247

247

247

247

247

Missing

0

0

1

1

0

0

0

0

0

Mean

4.36

4.05

3.98

4.78

3.98

3.33

4.32

4.46

4.74

Std. Error of Mean

0.105

0.106

0.114

0.109

0.092

0.086

0.108

0.097

0.107

Median

4

4

4

5

4

3

4

4

5

Mode

4

4

4

4

4

3

4

4

4

Std. Deviation

1.649

1.672

1.782

1.713

1.452

1.344

1.696

1.521

1.689

Variance

2.718

2.794

3.175

2.933

2.109

1.807

2.875

2.314

2.851

Skewness

-0.144

0.002

0.072

-0.483

0.035

0.771

-0.136

-0.324

-0.368

Std. Error of Skewness

0.155

0.155

0.155

0.155

0.155

0.155

0.155

0.155

0.155

Kurtosis

-0.626

-0.736

-0.869

-0.524

-0.371

0.426

-0.658

-0.417

-0.652

Std. Error of Kurtosis

0.309

0.309

0.309

0.309

0.309

0.309

0.309

0.309

0.309

Range

6

6

6

6

6

6

6

6

6

Minimum

1

1

1

1

1

1

1

1

1

Maximum

7

7

7

7

7

7

7

7

7

Research Tools Internal Consistency

The internal consistency of every questionnaire is a necessity and a key test is called the Cronbach’s alpha (α) measure and it is typically used to determine the reliability of Likert scales used in eliciting attitudes in research tools like questionnaires. SPSS provides a simple and quick way of assessing the reliability of the research tool. For the 9 items of the privacy perception part of the privacy perception questionnaire, it was found to have a Cronbach’s alpha (α) of 0.644 which is in between the good value (α > 0.7) and questionable (α < 0.6) by research standards (Gliem and Gliem, 2003). Table 4.12 shows the item total statistics for the 9 questionnaire items. For the complete analysis see appendix E.

Table 4.12 Total in item α statistics for the 9 item privacy perception part

Scale Mean if Item Deleted

Scale Variance if Item Deleted

Corrected Item-Total Correlation

Cronbach's Alpha if Item Deleted

Q1

33.62

39.893

.605

.544

Q2

33.91

39.172

.634

.535

Q3

33.99

40.246

.521

.562

Q4

33.19

40.538

.538

.560

Q5_R

33.97

54.175

-.049

.689

Q6

34.64

49.394

.215

.638

Q7

33.66

44.333

.354

.608

Q8_R

33.51

55.095

-.097

.701

Q9

33.22

47.408

.212

.642

Multivariate Analysis of Variance (MANOVA)

The mean differences in privacy perception in open plan offices is of primary concern in this study, if fact, it tests the main hypothesis of this study. Given that the survey instrument had a 9 item question set related to privacy with respondents from 3 different countries. The test of choice was the multivariate analysis of variance (MANOVA). The responses to the 9 item survey formed the dependent variables while the nationalities are the independent variables. Using SPSS, the MANOVA was carried out. The three nations where assigned dummy variables 1, 2, 3 corresponding to China, United States of America and Nigeria respectively.

The Wilks’ lambda statistics was used because of its widespread use and tolerance for example, Pillai's Trace result is considered unstable in non uniform samples as in this case but in any case all the MANOVA test statistics showed a significance p < 001. In SPSS, the Wilk’s lambda gives an exact value not subject to further treatment.

The MANOVA for the privacy perception scores by nationality showed effects for nationality, Wilks Lambda = 0.893, F (18, 468) = 4.32, p < 0.001 and confirms the main hypothesis that there will be differences in privacy perception in open plan offices across nations which we consider as cultural groupings.

A two way MANOVA for the data set as dependent variables was carried out with Nationality and Gender as the independent variables. Only Nationality showed a significant relationship where the Wilks’ Lambda = 0.788, F (18,462) = 3.26, p < 0.001. Then for Gender, Wilks’ Lambda = 0.95, F (9, 231) = 1.49, ns. While Gender * Nationality resulted in a Wilks’ Lambda = 0.92, F (18, 462) = 1.05, ns. This further shows that gender is not a specifically important consideration in the study of privacy perception in open plan office environments.

Subsequently, post hoc tests for the first MANOVA test for Nationality only showed the detailed interactions for Nation I - Nation J between the three nations for each of the items of the questionnaire and the complete results for the first MANOVA analyses is found in appendix F.

Exploratory Factor Analysis

An exploratory factor analysis (EFA) was embarked upon to uncover any underlying relationships within the privacy perception in open plan offices. In the COPE project study by Veitch et al. (2007), they found that the satisfaction with the office environment was related to three factors; privacy, ventilation and lighting and the privacy factor was made up of visual and acoustic privacy issues.

Generally, the aspects of privacy within open plan offices are grouped into audio and visual privacy issues explaining why noise and visual distractions are high. The factor analysis attempts to find out what other latent structures (Heppner and Heppner, 2009) exist within the privacy in open plan office construct. The missing values in the survey items were replaced by means in this analysis. Other applications of factor analysis include questionnaire design and validation, data reduction or management by reducing redundancy as represented by multicollinearity and singularity (Field, 2005).

The correlation matrix of the responses to questions of privacy perception part in second section of the research tool were examined and shown in Table 4.10 below. The correlation matrix was inspected for multicollinearity and singularity which indicates variables that are highly correlated (R > 0.9) and perfectly correlated respectively. These two issues affect the factor analysis in that through redundancy, they increase the difficulty in determining the unique contribution of each variable in the analysis. In SPSS, the R matrix’s determinant provides a quick test for multicollinearity and singularity. Field (2005) recommends that the determinant should be greater than 0.00001 and if less than the recommended figure, the correlation matrix is examined and one of the two variables that correlate too highly (R> 0.8) should be eliminated.

In the case of this research, no multicollinearity and singularity were detected the determinant of the R matrix was found to be 0.139, greater than the 0.00001 prescribed. This can be confirmed from an inspection of the correlation matrix in Table 4.13 below. For the full correlation, R matrix and determinant see appendix G.

Table 4.13: 9 item privacy perception part correlation coefficients

Q1

Q2

Q3

Q4

Q5_R

Q6

Q7

Q8_R

Q9

Correlation

Q1

-

Q2

.703

-

Q3

.537

.609

-

Q4

.463

.454

.322

-

Q5_R

-.019

-.006

-.019

.022

-

Q6

.198

.179

.102

.204

-.019

-

Q7

.298

.286

.302

.331

-.151

.156

-

Q8_R

-.072

-.057

.015

-.066

.140

-.098

-.089

-

Q9

.149

.180

.124

.322

-.175

.168

.256

-.183

-

The correlation coefficients are said to fluctuate with increases in sample size so, SPSS provides a test known as the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity. It is expected that result for the KMO should be greater than 0.5. According to Field (2005) KMO values within 0.5 and 0.7 are considered to be mediocre and within 0.7 to 0.8 as good, within 0.8 to 0.9 great while KMO values above 0.9 are excellent. The Bartlett’s test measures the null hypothesis that the privacy perception correlation matrix is an identity matrix for this sample as shown in Table 4. 11 the test is highly significant (p < 0.001) and the KMO sampling adequacy score was 0.78 which falls within the range of acceptability as shown in Table 4.14 below.

Table 4.14: KMO test of sampling adequacy

Kaiser-Meyer-Olkin Measure of Sampling Adequacy.

.780

Bartlett's Test of Sphericity

Approx. Chi-Square

474.234

df

36

Sig.

.000

The factor extraction was done using principal component analysis which is justifiable based on the normality criteria and values less than 0.4 where suppressed. The initial run of the factor analysis resulted in two factor groupings with the eigenvalues above 1 which is the usually recommended cut off point. The variable Q6 was suppressed while it had shown a significant variation in the MANOVA earlier carried out. The Questionnaire item Q6 was included to address the feelings of crowding in open plan offices especially as relates to workspace distances in open plan offices. The factor analysis was rerun and the number of factors specified to be three instead of the initial two.

The three factor solution seemed justifiable because an examination of the scree plot Figure 4.6 shows that the third factor hovered around the eigenvalue of 1 in fact, it was 0.977 which really approximates to 1.

After a Varimax rotation with Kaiser Normalization, the first factor accounted for 29.94% of variance while the second factor accounted for 16.37% of the variance and the third 12.70% of the variance. These three cumulatively accounted for 59.01% of the total variance. This is shown in Table 4.15.

Table 4. 15: The variance explained by the 3 factors

Component

Extraction Sums of Squared Loadings

Rotation Sums of Squared Loadings

Total

% of Variance

Cumulative %

Total

% of Variance

Cumulative %

1

2.981

33.117

33.117

2.694

29.937

29.937

2

1.353

15.039

48.156

1.473

16.372

46.309

3

.977

10.856

59.012

1.143

12.703

59.012

Extraction Method: Principal Component Analysis

Privacy Perception Factors

The first factor was mainly related to questions 2, 1, 3, 4 and 7 in that order and denoted as Q2, Q1, Q3, Q4, Q7 in Table 4.14. These questions relate to the issues of distraction by colleagues and office machines (both audio and visual distractions) as such, it is tagged “Distractions by Humans and Office Machines.” The next factor grouping was mainly related to questions 5 (reverse scored), 9 and 8 (also reverse scored) denoted as Q5_R, Q9 and Q8_R respectively. These questions relate to the matters of personal speech privacy and freedom to alter the aesthetics of the work space as such, it is then tagged “Privacy of speech and aesthetics.” The last factor, which relates uniquely to question 6 denoted as Q6 and then partly to Q4 and Q5_R concerns the enclosure by walls and the distances from colleagues more related to territoriality exhibited in humans. This is thus tagged “Envelopment and Crowding.

Table 4.16 Showing the rotated component matrix

Distractions by Humans and Office Machines

Privacy of speech and aesthetics

Envelopment and Crowding

Q2

.863

Q1

.833

Q3

.814

Q4

.583

.403

Q7

.459

.436

Q5_R

-.726

.452

Q9

.616

Q8_R

-.565

Q6

.764

Multiple Regression Analysis

To test the sub hypotheses proposed in this study, a regression analysis was carried out using the factor scores generated by the exploratory factor analysis. The stepwise method was used since it was primarily an exploratory study of the construct of privacy perception in open plan offices with the effect of culture as specified in the Hofstedes cultural values model. The stepwise regressions main advantage is that it can identify which predictors or independent variables accounted for the most variation overall in the dependent variable.

Thus, the three factors extracted from the exploratory factor analysis; Distractions by colleagues and Office Machines, Privacy of speech and aesthetics, Envelopment and crowding were utilized as dependent variables. For the sake of brevity the dependent variables were listed as Distractions_Human_Machines, Privacy_Speech_Aesthetics and Envelopment_Crowding respectively. The independent variables used includes all of the Hofstedes dimensions of cultural values namely; power distance (PDI), individualism (IDV), masculinity (MAS), uncertainty avoidance index (UAI), long term orientation (LTO), and the recently added indulgence (IVR).

The first iteration using the first factor – Distraction by human and office machines resulted in the masculinity (MAS) dimension producing the highest significant predictive ability as shown in Table 4.17 below. The Durbin – Watson statistics was also within the specified range of 1 to 3 (Field, 2005) meaning the likelihood of independent errors are minimal.

Table 4.17 Regression model summary for Distraction by humans and office machines

Model

R

R Square

Adjusted R Square

Std. Error of the Estimate

Change Statistics

Durbin-Watson

R Square Change

F Change

df1

df2

Sig. F Change

1

.200a

.040

.036

.98173729

.040

10.238

1

245

.002

1.832

a. Predictors: (Constant), MAS

b. Dependent Variable: Distractions_Human_Machines

The results of the stepwise regression for the first factor, showed that in the first step, the masculinity (MAS) predictor accounted for a significant amount of variance in the privacy perception factor related to distraction by colleagues and office machines (R2 = 0.04, F (1, 245) = 10.23, p < 0.01). This in a way confirms one of the sub hypotheses of the study. The coefficients and constant of the regression equation are shown in Table 4.18 below. The unstandardized coefficients and Beta values indicate a negative relationship between the feelings of Distraction by colleagues and office machines and the masculinity dimension (MAS) proposed by Hofstede.

Table 4.18 Regression coefficients for Distraction by colleagues and office machines

Model

Unstandardized Coefficients

Standardized Coefficients

T

Sig.

B

Std. Error

Beta

1

(Constant)

1.418

.448

3.168

.002

MAS

-.024

.007

-.200

-3.200

.002

a. Dependent Variable: Distractions_Human_Machines

The second iteration was for the second privacy perception factor - Privacy of speech and aesthetics. Here, two independent variables of power distance (PDI) and individualism (IDV) showed a significant predictive capability. The regression model summary is shown in Table 4.19 below.

Table 4.19 Regression model summary for Privacy of speech and aesthetics

Model

R

R Square

Adjusted R Square

Std. Error of the Estimate

Change Statistics

Durbin-Watson

R Square Change

F Change

df1

df2

Sig. F Change

1

.164a

.027

.023

.98851486

.027

6.750

1

245

.010

2

.217b

.047

.039

.98012130

.020

5.214

1

244

.023

1.785

a. Predictors: (Constant), PDI

b. Predictors: (Constant), PDI, IDV

c. Dependent Variable: Privacy_Speech_Aesthetics

This time around, on the first step of the second iteration the power distance dimension (PDI) was selected accounting for a significant variation in the Privacy for speech and aesthetics factor (R2 = 0.027, F (1, 245) = 6.75, p < 0.05). Then on the next step, the individualism dimension (IDV) was selected accounting for a significant amount of additional variance of privacy for speech and aesthetics factor (R2 = 0.047, R2 change (Δ R2) = 0.02, F (1, 244) = 5.21, p < 0.05). The Durbin – Watson statistics gave a value of 1.79 which is within the acceptable range.

Table 4.20 Regression coefficients for Privacy of speech and aesthetics

ModelpeUnstandardized CoefficientsraSig.daPDInsIDVns00a. Dependent Variable: Pr

Unstandardized CoefficientsraSig.daPDInsIDVns00a. Dependent Variable: Privacy_Speech_Aes

Standardized CoefficientststtaSig.ഴaBig

Std. Error

edBeta E

(Constant)

ed

TtaSig.ഴaBig

Std. Error

edBeta E

(Constant)

edMASns22a. Dependent Variabl

Sig.daPDInsIDVns00a. Dependent Variable: Privacy_Speech_Aesthetics0esUnstan

Big

Std. Error

edBeta E

(Constant)

edMASns22a. Dependent Variable: Distracti

Std. Error

edBeta E

(Constant)

edMASns22a. Dependent Variable: Distractions_Hu

Beta E

(Constant)

ed

1

(Constant)

edMASns22a. Dependent Variable: Distractions_Human_MachinesŀĈModelpe

-.612

.244

-2.510

.013

PDInsIDVns00a. Dependent Variable: Privacy_Speech_Aesthetics0esUnstandardized Coefficients

.009

.004

.164

2.598

.010

2

(Constant)

edMASns22a. Dependent Variable: Distractions_Human_MachinesŀĈModelpe

-11.042

4.574

-2.414

.017

PDInsIDVns00a. Dependent Variable: Privacy_Speech_Aesthetics0esUnstandardized Coefficients

.125

.051

2.244

2.457

.015

IDVns00a. Dependent Variable: Privacy_Speech_Aesthetics0esUnstandardized Coefficientsac--a. De

.064

.028

2.086

2.283

.023

a. Dependent Variable: Privacy_Speech_Aesthetics

The regression coefficients of the privacy for speech and aesthetics showed a positive relationship with the power distance dimension (PDI) and also with the individualism dimension (IDV) as shown in Table 4.20 above. Looking at the strengths of contributions, the power distance dimension (PDI) with t (245) = 2.46, p < 0.05 and individualism dimension (IDV) having t (244) = 2.28, p < 0.05; both being significant but with power distance dimension (PDI) having a slightly stronger contribution than the individualism dimension (IDV).

The third iteration was for the third privacy perception factor - Envelopment and Crowding and again, the masculinity (MAS) dimension showed the most significant predictive capability of the whole group of independent variables as shown in Table 4.19 below.

Table 4.21 Regression model summary for Envelopment and Crowding

Model

R

R Square

Adjusted R Square

Std. Error of the Estimate

Change Statistics

Durbin-Watson

R Square Change

F Change

df1

df2

Sig. F Change

1

.137a

.019

.015

.99260893

.019

4.677

1

245

.032

2.034

a. Predictors: (Constant), MAS

b. Dependent Variable: Envelopment_Crowding

In the last iteration, the masculinity dimension (MAS) was again found to account for a significant amount of variation in the Envelopment and Crowding factor (R2 = 0.019, F (1, 245) = 4.68, p < 0.05). The Durbin – Watson statistic of 2.034 was also within the acceptable range. The regression coefficients and related data are shown in Table 4.22 below.

Table 4.22 Regression coefficients for Envelopment and Crowding

ModelpeUnstandardized CoefficientsraSig.daPDInsIDVns00a. Dependent Variable: Pr

Unstandardized Coefficientsac--a. Dependent Variable: Envelopment_Crowdiness-ic

Standardized CoefficientststtaSig.ഴaBig

Std. Error

edBeta E

(Constant)

ed

TtaSig.ഴaBig

Std. Error

edBeta E

(Constant)

edMASns22a. Dependent Variabl

Sig.daPDInsIDVns00a. Dependent Variable: Privacy_Speech_Aesthetics0esUnstan

Big

Std. Error

edBeta E

(Constant)

edMASns22a. Dependent Variable: Distracti

Std. Error

edBeta E

(Constant)

edMASns22a. Dependent Variable: Distractions_Hu

Beta E

(Constant)

ed

1

(Constant)

edMASns22a. Dependent Variable: Distractions_Human_MachinesŀĈModelpe

.969

.453

2.141

.033

MASns22a. Dependent Variable: Distractions_Human_MachinesŀĈModelpeUnstandardized Coefficients

-.016

.008

-.137

-2.163

.032

a. Dependent Variable: Envelopment_Crowding

The coefficients of regression show that the masculinity dimension (MAS) showed a negative relationship with Envelopment and Crowding t (245) = -2.16, p < 0.05. All other multiple regression analysis not presented can be found in appendix H.

CHAPTER 5

CONCLUSION AND SUMMARY

This study set out to examine the role of culture in the perception of privacy in open plan office environments, employing ideas from a questionnaire based on satisfaction with environmental features (SEF) which was used by Veitch et al. (2007) and adapted to become a scenario survey of privacy perception across three nations of China, United States of America and Nigeria.

The influence of culture has been studied and known to provide explanations for human behavior and to aid study and research, one of the commonly used frameworks or models is the Hofstede dimensions of cultural values. The Hofstede framework as presented has six dimensions namely, power distance (PDI), individualism (IDV), masculinity (MAS), uncertainty avoidance index (UAI), long term orientation (LTO), and the recently added indulgence (IVR) for which all nations could have values.

After a design of the questionnaire and pilot testing, it was deployed as hardcopy and online versions in English for Nigeria and United States of America while that for China was translated to Mandarin. At the close of data collection, 247 responses where considered valid and the breakdown showed that there were 104 Chinese respondents, 75 Americans and 68 Nigerians. The sample size was not as large as expected but tests of sampling adequacy showed that the sampling was adequate with the KMO > 0.7.

The general statistics and in formation obtained from the first section of the questionnaire was analyzed and it showed among other things the gender distribution of the respondents with 143 males 104 female.

General Information

The general perception of the popularity of the open plan office despite the general dislike for the open plan office was also highlighted by the answers to the general questions. The data showed that 73% of all respondents worked in open plan offices which cover landscaped offices and cubicles with low walls. 26% of the respondents worked in some form of cellular office which includes the traditional office or offices with thin walls or transparent walls made for example with glass. While 1% worked from home i.e. they had no specific workspace at work.

Subsequently, the choice of the ideal office type was also elicited. Then, the reasons for the choice of a particular office type were also elicited. The results showed that if the open plan office to the traditional office is thought of as a continuum with the cubicle with low walled partitions and then the thin walled or transparent walled offices as intermediaries, the main reasons for the choice of each office along the continuum shows a kind of an overlapping stepped characteristic. This is represented by figure 5.1 below.

Reasons

Professional status

Personal privacy

Minimal distraction

Interaction

Open plan/ Landscape

Cubicle/Low Partitions

Thin/See through walled

Cellular/ Traditional

Office Type

Figure 5.1 Stepped characteristics of the reasons for the choice of office type.

Preliminary Analysis and Main Hypothesis Test

The preliminary analysis of the data using the descriptive statistics minimized the chances of error and helped affirm the assumption of normality and the measure of variation also called heteroscedasticity. The Cronbach’s alpha value for the 9 item privacy perception survey part was α = 6.44 which was acceptable but less that the preferred value of 0.7.

The task of finding out if there were significant variations in the 9 item responses across nations was checked using a multivariate analysis of variance (MANOVA). First a one way MANOVA was carried out for the three nations and it showed significant variation at ρ < 0.05 then due to some research reports (Anjum et al., 2004; Kaya and Weber, 2003; Newell, 1998; Yildrim et al., 2007) suggesting that gender could affect privacy response and desire or privacy issues in general, another two way MANOVA for gender and nationality was carried out. It still showed that only nationality accounted for any significant variations in the means of the 9 item privacy perception responses.

The MANOVA results confirmed the variation of privacy perception across national and hence cultural lines thus, confirming the main hypothesis of the thesis. The 9 item privacy perception responses were then treated in an exploratory factor analysis (EFA) and it resulted in a 3 factor solution for the privacy perception construct in open plan offices.

Three Privacy Perception Factors

The exploratory factor analysis showed that the idea of privacy can be approached from three aspects discussed below.

The first is concerned with distractions by colleagues and an office machines. This was called Distraction by Humans and Machines factor. This was the first factor extracted during the exploratory factor analysis (EFA). This factor is related to the general idea of audio and visual distractions in open plan offices caused by the general ambient noise, specific conversation or noise from office and personal devices like copiers and phones.

The second factor to be considered is the confidentiality provided by the workspace for personal speech with a client, colleague, family or acquaintance on the phone for example. Secondly, the idea of personalization was also included in this factor as humans and animals tend to exhibit territoriality with personal and aesthetic additions to the workspace as such, pictures, flowers and trophies tend to mark workspaces. These issues were called the Privacy for Speech and Aesthetics factor.

Lastly, the feelings of crowding and envelopment formed another factor. This third factor was called Envelopment and crowding factor. Some studies show that person to person proximity measures differ from person to person and especially within cultures (Hall, 1966). Hence, the general idea of intruding in on someone’s personal space. The general understanding is that with an increase in enclosure as represented by wall or partition height an increase in privacy is perceived (Oldham et al., 19911; Yildrim et al., 2007).

Sub Hypothesis Tests

The sub hypotheses were tested using a stepwise multiple regression, of the three privacy perception factors as the dependent variables or criterion while the 6 dimensions of Hofstedes cultural values as the independent variables or predictors. The results obtained showed the following:

For the first factor, related to perception of distractions from colleagues and office machines was negatively related to masculinity dimensions (MAS) which meant as the femininity rises, the perception of distraction from colleagues and office machines also rises. The model fit was low at 4%. The nature of femininity or low scores on the masculinity and its relation to modesty may provide an explanation for the increased feelings of distraction from colleagues and even office machines.

Secondly, for the privacy for private speech and aesthetics, two predictors were entered into the model. The dimensions of power distance (PDI) was entered first with a model fit of around 3% and then the individualism dimension (IDV) with a model fit of 5% indicating a change of just 2% in the fit statistics. It is not surprising that the power distance (PDI) and individualism dimensions (IDV) affected the perception of speech privacy and aesthetics the first part which is related to information privacy and the second more related to physical privacy or even more of territoriality (Leino- Kilpi et al., 2001).

Thirdly, the last factor related to envelopment and crowding, the regression statistics showed the masculinity dimension (MAS) accounted for a significant variation and was yet negatively related to the envelopment and crowding dimension. The model fit statistics revealed that the masculinity (MAS) dimension only explained 2% of the variation in the third factor. This shows that an increase in femininity influenced the perception of crowding and envelopment a factor of privacy perception.

The regression results show that the three dimensions of cultural values as represented by masculinity (MAS), power distance (PDI) and individualism (IDV) had an impact on privacy perception in open plan offices. While the uncertainty avoidance index (UAI), long term orientation (LTO) and the newly added indulgence (IVR) showed no significant relationships to the concept of privacy perception in open plan offices.

Limitations of the Study

This study was an attempt at a pan cultural but opportunistic study (Sekaran, 1983) of privacy perception in open plan offices. The nature of the study resulted in many constraints. The number of respondents in general could be considered low and even the number of nations is also low as Meeuwesen et al., (2009) considered a 10 nation study to be low as related to the number of nations.

The Hofstedes cultural values scores for Nigeria in particular were taken from the general West African scores except for the long term orientation (LTO) and individualism (IDV) scores. This could be a source of inconsistencies.

Another important issue is the effect of organizational culture; this seems to be one of the strengths of Hofstedes initial study which was carried out within staff of a single multinational corporation and even the type of work as certain tasks with low complexity allow for some measure of distraction while others do not (Oldham et al., 1991). This study did not discriminate between workers and their work space types this may have included persons with little or no experience of open plan office environments. The total time spent working at the work station could also provide some bias in the perception of privacy.

It could be inferred from the results of this study, that the typical open plan offices will no personal enclosures, poor noise regulation and perceived or real crowding will be less well received in countries with very high scores in Power distance dimension (PDI), Individualism (IDV) and low scores for the Masculinity dimension (MAS) which translates to very high femininity. Given that the workplace is an organizational resource (Vischer, 2006) its effective use may be influenced if cultural factors are not taken into consideration in the design and assignment of work spaces.

Recommendations for Further Research.

The number of nations in this study could be expanded to capture more effects and make it a truly pan cultural study. The ranking algorithm could be employed for the choice of nations to be used in hypothesis testing as proposed by Sivakumar and Nakata (2001) to find a minimal but effective group of nations for this and similar studies. This in effect suggest a move from opportunistic approaches to this study except for a truly pan cultural approach where the more nations the better.

The predictive capability of the Hofstedes culture dimension showed a very low model fit in this study as such, further studies should examine the nature of work i.e. the task complexity as it affects the perception of distractions both of the audio and visual variety (Banbury and Berry, 2005). Subsequent studies could also be carried out within office workers of a single industry for example design Engineers, Architects or Accountants etc. to minimize the impact of differences in task complexity in the perception of privacy in open plan environments.

As stated earlier in the literature review, Newell (1998) noted that there seemed to be a fit of cross-cultural desire of privacy to the universal value scheme proposed by Schwartz and Bilsky (1987) namely; self-direction, universalism, benevolence, tradition, conformity, security, power, achievement, hedonism and stimulation. This will be a framework with which to approach the culture question in the perception of privacy in open plan offices in subsequent studies.

The role of organizational culture could be studied as corporate culture affects the workplace also through regulations and controls, motivation and even the design of the workplace.

Writing Services

Essay Writing
Service

Find out how the very best essay writing service can help you accomplish more and achieve higher marks today.

Assignment Writing Service

From complicated assignments to tricky tasks, our experts can tackle virtually any question thrown at them.

Dissertation Writing Service

A dissertation (also known as a thesis or research project) is probably the most important piece of work for any student! From full dissertations to individual chapters, we’re on hand to support you.

Coursework Writing Service

Our expert qualified writers can help you get your coursework right first time, every time.

Dissertation Proposal Service

The first step to completing a dissertation is to create a proposal that talks about what you wish to do. Our experts can design suitable methodologies - perfect to help you get started with a dissertation.

Report Writing
Service

Reports for any audience. Perfectly structured, professionally written, and tailored to suit your exact requirements.

Essay Skeleton Answer Service

If you’re just looking for some help to get started on an essay, our outline service provides you with a perfect essay plan.

Marking & Proofreading Service

Not sure if your work is hitting the mark? Struggling to get feedback from your lecturer? Our premium marking service was created just for you - get the feedback you deserve now.

Exam Revision
Service

Exams can be one of the most stressful experiences you’ll ever have! Revision is key, and we’re here to help. With custom created revision notes and exam answers, you’ll never feel underprepared again.