Measuring Diversity Practice And Developing Inclusion Commerce Essay

Published:

The issues of workforce diversity and inclusion may be one of the biggest challenges facing organisations over the next decade. The revolution in technology has led to the workplace becoming more global, and changes in the political landscape have led to the erosion of national boundaries and the emergence of developing markets. For organisations, the pool of potential stakeholders has transformed, including customers, suppliers and employees. In terms of studies in the field, the academics appear to be slow off the mark. The key issues being discussed in the literature are:

How shall we conceptualise diversity and what are its core features? Should diversity be defined in terms of socio-demographic or ethnic categories only, or does diversity include a wide range of values, attitudes and skills in which people working together at various levels of an organisation may differ?

What is the difference between diversity and equal opportunity? Moving this line of reasoning forwards, what are the differences between diversity, equal opportunities and inclusion (see for example the 'diversity model' in Ashkansasy et al, 2002)?

Lady using a tablet
Lady using a tablet

Professional

Essay Writers

Lady Using Tablet

Get your grade
or your money back

using our Essay Writing Service!

Essay Writing Service

What are the salient indicators of a diverse workforce? How can we measure diversity among the workforce? How can we measure inclusion?

How can we know if diversity and inclusion is improving in our organisation? How do we know if our diversity and inclusion initiatives are working for us?

According to April & Shockley (2007), understanding of diversity and workplace practice has undergone significant philosophical shifts over time. During the early to mid-20th century, the focus of diversity was initially driven at a country level, driven on two overriding philosophies: the women's rights and civil rights philosophies. The women's rights philosophy focused particularly on eliminating sexism, and the civil rights philosophy sought to end discrimination and racism, and was mainly focused at minorities. Citizens, civil society, the private sector and governments thereafter found themselves required to comply with legal requirements that guaranteed citizens certain rights.

There are occasions where the history of a country, or a specific context, dictates the need to focus more narrowly on diversity through compliance initiatives - such as in the case of post-Apartheid South Africa, where in trying to redress the imbalance of their Apartheid past, the South African government and its people have legislated organisational focus on previously-disadvantaged groups (i.e., specifically Black people which includes the Indian, Asian and Coloured populations, and all women); this was also true in Malaysia when the Bumiputra's came into power. New Zealand followed, and is following, a similar path with regards to the Maoris, and America's affirmative action focus produced leaders such as Colin Powell.

Such compliance initiatives tend to build resistance. It does not take very long within workplaces, however, before organisations start complaining about the cost that such compliance is adding to their operational costs. Additionally, many internal employees start complaining - those not being affirmed feel left out, claim reverse discrimination, engage in confused collaborative-disingenuous relationships with those being affirmed, and some even start psychologically withdrawing from their companies (translating in less effort); those being affirmed feel that their credibility is being put under question, are having to be many times as good as others to feel credible, and are carrying the burden of their entire minority group (i.e., feeling that if they fail, their employers and employees and those not being affirmed will write-off their entire minority group).

In the last decade of the 20th century, with growing globalisation and cross-regional cooperation and interaction, the shift in focus moved from a legalistic approach, to one of 'valuing diversity'. The American Psychological Association (APA, 2006) claims that 'valuing diversity is what institutions and members of a community do to acknowledge the benefits of their differences and similarities. They intentionally work to build sustainable relationships among people and institutions with diverse membership. A community that values diversity ensures that institutions provide equal treatment and access to resources and decisions for all community members regardless of race, ethnicity, sexual orientation, and physical disability. This socially responsible philosophy was based on a view of a more international (in outlook, at least) organisation, practising good corporate citizenship, with its leadership/management supposedly wanting their organisations to act in ways that benefit all of society. A few forward-thinking organisations understood that in sectors and industries where many products and services are undifferentiated between players in the same/similar industries, responsible behaviour that is contextually relevant could lead to competitive advantage (Cascio, 1998; Johnston, 1991).

Lady using a tablet
Lady using a tablet

Comprehensive

Writing Services

Lady Using Tablet

Plagiarism-free
Always on Time

Marked to Standard

Order Now

The 'valuing diversity' philosophy expanded the differences that mattered from initially being about ethnicity, gender, age and disability (observable differences) to include some underlying attributes such as values, socio-economic background, education, thinking styles, skills and knowledge, personality, and so on (Milliken & Martins, 1996; Hayles & Mendez-Russell, 1997). Companies such as Pillsbury Co. then defined diversity as "all the ways in which we differ"; Monsanto Agricultural Group defined diversity as "all the ways in which people differ and the effect of those differences on our thinking and behaviour"; and Medtronic Inc. defined it as "recognition, acceptance, and respect for individual differences and the awareness that these differences affect the ways employees work and interact with each other. Each person is influenced by characteristics such as age, gender, nationality, physical ability, race, sexual orientation, culture, values, attitudes and behavioural style which make her or him uniquely different from others".

Diversity was assumed to affect an organisation's performance by expanding its ability to serve a broader customer base, acquire more diverse and local talent pools, and procure resources and respond to environmental changes (April & Shockley, 2007; Roberson & Park, 2004). There was recognition that diversity-related initiatives could improve an organisation's ability to communicate its value in capital markets and could provide for the provision of superior services, because it enabled organisations to better understand customers (Wentling & Palma-Rivas, 2000), enabled organisations to tap into niche markets (Mueller, 1998), enhanced flexibility through diversifying market segments (Fleury, 1999), and gave them capability to respond to change more quickly (Adler, 1997; Jackson et al, 1992).

It is understood that if an organisation does not do well in 'valuing diversity', it can actually cause harm to the organisation's reputation externally, and internally the harm would manifest itself through inefficient communication, high interpersonal conflict and increased employee turnover (Bennett-Alexander, 2000).

Inclusion

At the start of the 21century, much was written about diversity relating to the context of organisations, e.g., recruitment, selection, placement, succession planning, performance management and rewards (Cascio, 1998), the impact of the culture of diversity on the quality of groups and teams (Cox, 1994); while still others claimed empirical investigations (Thomas, 2004). However, little can be found on the next philosophical evolution, that of 'inclusion'. Pearpoint (1990), Barak (2005) and Burnett & Kettleborough (2007) are some of the few voices in the field that is trying to move the debate from one of diversity to one of inclusion. A couple of papers have attempted to review the diversity debate and associated research findings (see Milliken & Martins, 1996 and Williams & O'Reilly, 1998), while others have looked at tangential areas. Ashkanasy et al (2002) for example, have looked at the issue of emotions linked with inclusion in organisations, while Jackson et al (2003) try to identify new areas for research.

To fully embrace the 'inclusion' philosophy organisations were encouraged to move beyond focusing on "some", and move beyond the rhetoric of how "all" were benefiting as a by-product of the focus on "some" (under the previous philosophy). In practice, this meant that managers/leaders had to not only develop an environment which increased the motivation, satisfaction and commitment of diverse people (Subhash, 2003), but furthermore it had to be an equitable work environment where no group had an advantage or disadvantage (Torres & Bruxelles, 1992). Additionally, inclusive workplace communities utilise the talents of people who would otherwise be discarded and written off - sometimes unconsciously excluded by dominant paradigms or rank.

Companies around the globe started defining inclusion as, for example, SABMiller "… we endeavour to create an inclusive culture where all employees feel appreciated for their uniqueness and their contributions are respected"; or Novartis "…it takes inclusive leadership, in spirit and in practice, to secure the business strategy's success. Only by motivating positive team dynamics and productivity, leveraging the differences, and bringing about the best possible contribution from each individual, can we deliver on the promise of diversity and inclusion". The shift is shown in Figure 1, and encourages management/leadership to provide real resources and executive support (Hayes, 1999; Jackson et al, 1992; Harris, 1989) for diversity efforts that includes everyone and has bottom-line and competitive advantage effects.

Figure 1: Difference between Diversity and Inclusion Philosophies

Lady using a tablet
Lady using a tablet

This Essay is

a Student's Work

Lady Using Tablet

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

Examples of our work

The diversity business rationale (April & Shockley, 2007; Cox & Blake, 1991; Hayles & Mendez-Russell, 1997; Davidson, 2002; Kossek et al, 2004; Allen & Montgomery, 2001) can be summarised as:

enhanced skill in entering untapped markets; understanding the needs of global customers, by engaging them through local knowledge, local relationships and local employees;

innovation and creativity flowing from new perspectives;

increased organisational sensing capability, as well as the ability to adapt to changing environments;

attracting, retaining and fully developing "all" staff, not just "some";

maximising value and productivity by seeking to understand the link between diversity and organisational performance;

being regarded as an employer of choice, because "all" people are valued and the organisation represents the demographics of the regions in which they operate;

development of a greater pool of leadership capacity; and

growth and profitability, resulting from the embedded resilience resulting from the "requisite variety" in the organisation.

Pless & Maak (2004) argue for a culture of inclusion founded on reciprocal understanding, standpoint plurality and mutual enabling, trust and integrity, while Miller (1998:152) suggests that 'building an inclusive organisation requires a serious commitment to fundamental change in the structures, behaviours, operating procedures, human resource systems, formal and informal reward systems, leadership practices, competency requirements and culture of the organisation.' Despite the increase in diversity policies and diversity awareness campaigns in the workplace, Allen & Montgomery (2001) found that they often fail because they typically use a scattershot approach instead of a planned approach to organisational change. They feel that diversity and inclusion is predominantly a total culture change. Change at individual, group and organisational- levels. This is mirrored by Thomas & Ely (1996) who maintain that effective diversity and inclusion requires a fundamental change in the attitudes and behaviours of an organisations leadership. Wilson (2000), in an in-depth study of three organisations, demonstrated that organisational behaviour and workplace culture are critical factors in delivering a diverse and inclusive organisation. Hopkins et al (2001) call for inclusive climates, which can be achieved through training (Thomas, 1994), organisational communication (Anfuso, 1995) and corporate values and mission statements (Milburn, 1997). Kochan et al (2003) argue that the positive and/or negative impact of diversity on outcomes such as performance hinges on organisational practices as well as on the culture and business strategy of the organisation, but that it has no direct impact itself. That is, diversity operates through various group processes such as communication, negotiation or conflict resolution, and our personal perception and stereotypes in turn impact on these. Such an indirect effect is difficult therefore to measure and any impetus to improve diversity becomes difficult to evaluate.

Thomas & Ely (1996:88) suggest that 'diversity should be understood as the varied perspectives and approaches to work that members of different identity groups bring' and make the case that diversity within a workforce has the potential to increase organisational effectiveness, hence differentiating it from discrimination and equality issues which have moral or legal ends and are often imposed from the 'outside'.

As a result, more recently, the 'inclusion' philosophy has been expanded (Figure 2), beyond the interpersonal/relationship team level, to: (1) behavioural manifestations of neurological (cognitive) and biological (emotional) circuitry - termed, "self leadership" - which challenges individuals to manage/lead themselves, manage their own prejudices and stereotypes, seek awareness into the ways in which they subtly damage the self-confidence and self-esteem of those with whom they work and live; to understand the influence of their own intentions on lived-behaviour; to deconstruct the way in which they negotiate their identities as individuals in networks of power; and to constantly expose their own views, and their view of others, to challenge and debate (April & Shockley, 2007; Schoem et al, 1995); and (2) casually and quantitatively link all of the inclusion insights to performance measures.

Currently around the world, there is scant evidence of organisations engaged in such work, and even less academic energy on the topic. There seems to be a myth operating that suggests that the outcomes created by inclusion processes defy measurement or can only be measured in the long-term. In a sense, the underlying fundamental belief is that creating an effective and inclusive work environment is something of a complex, soft but mysterious art form (Hubbard, 2001). While peers in other organisational areas are focusing on metrics which reflect their contribution such as sales, reduced costs, market share, return on investment, profits, income, expenses, and so on, those implementing the inclusion processes tend to limit its contribution to increased awareness, improved feelings and cultural awareness.

It is a real missed opportunity, because the result is that inclusion is not taken seriously. Fewer managers support it in actual practice and the initiative experiences less follow-through than other business initiatives. Baytos (1995: 21) claims "While I don't question the sincerity of the interest in identifying the value of diversity interventions, I sometimes wonder if companies are attempting to apply a tougher standard of evaluation for diversity processes than they do for other ongoing activities". Indeed, Ferdman & Davidson (2002) go so far as to claim that choosing not to engage in dialogue about inclusion in almost any modern organisation is just plain dumb!

Figure 2: Inclusion: Three Levels of Organisational Focus

There have not been many published, rigorous attempts at trying to measure inclusive

workplaces. The Australian Government, for instance, offers a diversity management feedback system free of charge (Bean et al, 2001) based on survey tools developed by Bean & Dillon (2000) which gives an organisation feedback at the individual, workgroup and organisation level on five diversity climate issues: respect, equality, conflict, discrimination and feelings about diversity. The organisation is given an overall score of poor, below average, average to good, or excellent according to how individuals score the 30 items in the survey.

Alternatives are offered by the National Extension Diversity Centre, a division of R Thomas Consulting in the USA, or Pearn Kandola Occupational Psychologists, an HR consultancy in the UK. Most measures of diversity tend to be raw numbers, quotas, or proportions of the workforce demonstrating such areas of difference looking at diversity as an input target. Some organisations monitor issues such as turnover amongst diverse groups, complaints made on the grounds of fairness and equality, recruitment costs and exit interview analysis to gain some measure of output from their diversity policies and practices. These largely address the failures of the diversity practice rather than areas of success, and can be widely impacted by external factors, such as the state of the labour market generally.

It is for this reason that the InclusionIndexTM was developed - a 10 dimension measurement framework, which gives real insight into inclusive environments for organisations on the following factors: senior managers, immediate manager, values, recruitment, promotion/progression & development, fitting in, bullying/harassment, dialogue, organisational belonging, and emotional well-being. The index gives organisations an indication of where their diversity and inclusion practices are failing, and where resources and effort are need to be applied to achieve the necessary change.

Figure 3: InclusionIndexTM Framework

Any measure therefore of whether or not a development initiative is improving inclusion and diversity can be obtained by running the index in the organisation prior to any initiative taking place, and then again some time afterwards. This has been tried by a large UK banking operation (UKBankPLC) with a two year gap to allow for the inclusion activity to take affect. The results are discussed as part of the findings presented in this paper.

Methodology

Each item set was statistically analysed using univariate and graphical data analysis of the individual items to detect distributional problems. In addition, the item means were checked for high levels of acceptance or rejection so to detect working problems or a response bias. The boxplots were checked for distributional anomalies.

Correlational analysis was used to investigate whether the items were sufficiently intercorrelated as indicataed by the KMO index. A desireable value would be >.80. In addition the individual MSA values were also inspected to identify individual items that are poorly related with the other items in a set and should therefore be removed or altered.

A Principal Component Analysis (PCA) was used to test whether each set of items is unideminsional meaning that all the items can be more or less represented by one dimension. Items with low loadings on the first factor were identified as problematic. The correlation matrix was based on a maximum number of data for each correlation (ie the 'pairwise' option for handling missing data).

An analysis of the internal consistency of each item set was carried out including the calculation of Cronbach's alpha as a measure of reliability of the sumscale. Items that did not contribute to the internal consistency of the additive scale were identified for exclusion. This analysis was based on complete data records (ie the 'listwise' option for handling missing data).

The sample for this analysis was N=3679 and included only those participants of the survey with no more than 17 missing values on the 53 inclusion-index items. For the majority of items the percentage of missing values was well below 10% which was considered to be a very good result. This paper reports the third iteration of the inclusion index, with previous iterations requiring substantial reworking following their validity studies.

The index was then implemented in UKBankPLC in 2005 (n=3,735) and again in 2007 (n=3,909) representing response rates of over 50% of the organisation on both occasions, and the index results were statistically analysed using crosstabulations and ANOVA.

Results

A summary of the statistical analysis is presented in table 1. The results show that the frequency distributions of the items are within normal range and show no grave anomalies. Each set of items is sufficiently well intercorrelated (KMO index >.80). Each set of items can be regarded as mainly unidimensional, which is a prerequisite for the calculation of a sumscore. Each set of items is sufficiently homogenous. The reliabilities of the inclusion-index factors are all high to very high. The reliabilities of the organisational outcome scales are excellent. The principal component analysis revealed a small second factor for values, recruitment, and promotion/progression & development, which are noted but do not cause grave concern with regard to the overall validity of the instrument.

Item Set

Univariate Analysis

Correlational Analysis

Principal Component Analysis

Internal Consistency

Senior Managers

(7 items)

No unusual findings

Degree of correlation is very good: KMO=.88

1 strong general factor confirming unidimensionality

% of total Var:49%

Reliability of sumscale is very good:

Alpha=.81

Immediate Manager

(6 items)

No unusual findings

High level of endorsement for several items

Degree of correlation among the items close to excellent:

KMO=.89

1 strong general factor confirming unidimensionality

% of total Var:63%

Reliability of sumscale is high:

Alpha = .88

Values

(10 items)

No unusual findings

Peaked distribution for Q19

Degree of correlation among the items close to excellent:

KMO=.89

2 factors emerged:

a strong 1st factor (45%) and a minor 2nd factor (10%)

Reliability of sumscale is high:

Alpha=.86

Recruitment

(6 items)

No unusual findings

Degree of correlation is good:

KMO=.80

2 factors emerged:

a strong 1st factor (54%) and a small 2nd factor (17%)

Reliability of sumscale is very good:

Alpha=.83

Promotion, progression & Devt

(7 items)

No unusual findings

Degree of correlation is good:

KMO=.80

2 factors emerged:

a strong 1st factor (47%) and a small 2nd factor (18%)

Reliability of sumscale is very good:

Alpha = .80

Fitting In

(6 items)

No unusual findings

High level of endorsement for several items as well as peaked distributions

Degree of correlation is very good:

KMO=.83

1 strong general factor confirming unidimensionality

% of total Var:49%

Reliability of sumscale is good:

Alpha=.79

Bullying & Harrassment

(5 items)

No unusual findings

Some items had high means and skewed distributions as to be expected

Degree of correlation is good:

KMO=.83

1 strong 1st factor confirming unidimensionality

% of total Var: 68%

Reliability of sumscale is very high:

Alpha=.88

Dialogue

(6 items)

No unusual findings

Degree of correlation is very good:

KMO=.86

1 strong general factor confirming unidimensionality

% of total Var: 59%

Reliability of sumscale is very high:

Alpha=.86

Org'nal Belonging

(6 items)

No unusual findings

Degree of correlation is very good:

KMO=.88

1 strong general factor confirming unidimensionality:

% of total Var:70%

Reliability of sumscale is excellent:

Alpha=91

Emotional well-being (+ emotions)

(7 items)

No unusual findings

Degree of correlation is excellent:

KMO=.91

1 strong general factor confirming unidimensionality:

% of total Var: 64%

Reliability of sumscale is excellent:

Alpha=.90

Emotional well-being

(- emotions)

(7 items)

No unusual findings

Degree of correlation is very good:

KMO-.85

1 strong general factor confirming unidimensionality:

% of total Var: 59%

Reliability of sumscale is very high:

Alpha=.88

Table 1: Summary of findings validating the Inclusion Index measurement inventory.

Each item set will now be discussed in turn.

Senior Managers

The descriptive statistics and the boxplots of the items indicate no problems regarding the empirical range of the scores, the means of the items or the shape of the frequency distributions of the items, which are only slightly skewed towards agreeing. There were also few outliers. The items are sufficiently intercorrelated, with a very good KMO value of 0.86. A PCA resulted in only one common factor explaining 49% of the total variance. The loadings on this factor were all in excess of .50. Hence it can be concluded that this set of items is unidimensional, and therefore can be well represented by a sumscale computed by including all items. The internal consistency of the Senior Managers Scale is very good (Cronbach alpha: .81) with important contributions from each item.

In UKBankPLC, it was clear that the senior management support of the inclusion initiative was appreciated, as while senior managers scored 75% in the first run of the index, this increased to 81% in 2007. While overall perceptions of the senior leadership have improved, female colleagues, those of Islamic faith and part-time colleagues were significantly less positive than others (p<0.05).

Immediate Manager

The descriptive statistics show that all item means are above 3, and the medians are 4, suggesting overall a high level of satisfaction with their immediate managers. As a consequence, the frequency of distributions of the rating scales are somewhat skewed. This is no cause for concern as the scores still span the whole rating scales. Furthermore, there were only very few outliers and no extreme cases. All the items are highly intercorrelated, with a KMO value of 0.89. A PCA confirmed that this set of items is unidimensional as only one dominant factor emerged from the analysis explaining 63% of the total variance. All items had a substantial loading on this factor. The analysis of the internal consistency of a sumscale of these items revealed a very high value of 0.88 for Cronbach's alpha confirming that a sumscale can be computed for Immediate Manager including all the items tested.

In UKBankPLC this area was clearly discriminated in the results between the two time periods. When the data was cut against those who had and those who had not received diversity training, the scores for immediate managers either went up in a department or down. Where they went up, the managers were clearly supporting the practice that was being promoted through the training. Where the scores decreased, the organisation needs to pay attention to management practice, as it is most likely that the training has caused disillusionment as the employees are not seeing the rhetoric applied in reality and hence are scoring their manager lower than before their awareness was raised.

Values

The descriptive statistics showed somewhat elevated item means, and for two question also relatively high skewness values. Notably one question has a narrow frequency distribution with a vast majority of participants opting for the answer 'strongly agree'. All items are substantially intercorrelated with a KMO value of 0.89, and the MSA value indicates that no items should be removed from this set. The outcome of the PCA revealed a strong 1st factor (45%) and a minor 2nd factor. This second factor was mainly made up on one question as its loading on this factor was higher than on the 1st common factor. This question and one other contributed with negative loadings to the 2nd factor, which is difficult to interpret. However, the wording of the question item in consideration does ask the respondents to evaluate diversity in relations to the future, which could trigger a tendency to simply agree with the item. This question did also not contribute sufficiently to the sumscale, and hence could be removed altogether. Regardless of this question, the analysis of internal consistency showed that the items produced a Cronbach's alpha of 0.86, suggesting that the other items form an excellent sumscale.

The scores for the values at UKBankPLC increased between the two time periods, but worklife and flexible working perceptions had not changed.

Recruitment

The descriptive statistics showed normal values on all key indices and the boxplots revealed no distributional anomalies. All items are sufficiently intercorrelated with each other with a KMO value of 0.80. A PCA confirms that this set of items is essentially unidimensional as the 1st factor explained around 54% of the total variance, although there was a small 2nd factor composed of 2 questions only. These two questions are similar as they relate to 'diverse background' and 'diverse employee base'. Since both items also have high loadings on the 1st factor they do not cause a problem regarding the unidimensionality of this set of items. The analysis of internal consistency showed that the items produced a Cronbach's alpha of 0.83, suggesting that a sumscale can be computed including all the items tested. At some point in the future, though, the wording of the two questions generating the 2nd factor might be considered in order to achieve an even clearer result, but this is not necessary at this stage.

At UKBankPLC this was an area which was of concern for some departments, particularly when analysed on work-basis. Part-timers felt recruitment practices were less fair than full-timers (p<0.05), suggesting that there is still more work to be done in this area, either in reviewing the practice, or managing expectations if analysis of recruitment outcomes suggest that there is no bias or skew.

Promotion, Progression and Development

The values of all the relevant descriptive statistical indices fell into the normal range, so there are no critical issues regarding the frequency distributions of these items. The boxplots indicated a symmetrical shape of the frequency distributions without any extreme cases and very few outliers. There is sufficient spread in the scores across the ratings, and the magnitude of the correlations among these items is very good with a KMO value of 0.80. The PCA revealed only one strong factor explaining 47% of the total variance. Each item had a substantial loading on this 1st factor (>.50). There was however a small 2nd factor explaining an additional 18% of the variance, generated mainly from two questions relating to inquiring about issues relating to 'diversity and inclusion'. These two questions did however load sufficiently on the 1st factor to consider the item set to be unidimensional, however it would be worth monitoring and revisiting the two questions loading the 2nd factor in due course. The analysis of internal consistency produced a Cronbach's alpha of 0.80, and hence a sumscale can be constructed using all the items within this set.

At UKBankPLC this was the area that highlights most concern in terms of differences between the two time periods. While promotion scores increased overall, this was not the case for minority groups, whether the group was defined by gender, race, or work-basis (p<0.05). This could be because a raised awareness has enhanced individual expectations and the organisation has not yet managed to meet these aspirations, or it could be that employees have become aware of an area where practice could improve further. Further analysis is needed before conclusions can be drawn, but the index has clearly identified that this is an area for attention.

Fitting In

In this item set, three questions had means around 4 (ie strongly agree) combined with a peaked frequency distribution of the scores. This suggests that many participants throughout the sample rated their degree of 'fitting in' as similarly high. This does not create a problem since the correlational analysis showed that the items overall were well intercorrelated with a KMO value of 0.83. This suggests there is no problem with the wording of the three questions scoring highly. The PCA extracted only one general factor from each correlation matrix explaining 49% of the total variance. This confirms that there is just one dimension underlying the correlation matrix of these items. An analysis of the internal consistency of the sumscale of these items produced a Cronbach's alpha value of 0.79, and hence all items can be computed in the sumscale calculation. In order to raise the internal consistency of this scale further, an additional question should be added to the set.

In UKBankPLC, the sense of fitting in increased over the time period being studied, particularly amongst those who had received the diversity training (p<0.05).

Bullying and Harrassment

The descriptive statistics show high means and a skewed frequency distribution for two of the items. The wording of these questions refers to the respondent being 'threatened' or 'bullied' and hence it is to be expected that relatively few respondents will answer at the top end of the scale. The items were well intercorrelated with a KMO value of 0.83, and the individual MSA values suggest that no items should be removed from the set. The PCA produced only one strong general factor explaining around 68% of the total variance. All items had substantial loadings on this general factor confirming that the items can indeed be represented by a unidimensional scale. The internal consistency of a sumscale of these items revealed a Cronbach's alpha value of 0.89 and hence no revisions are necessary to this scale and all items can be considered in the sumscale.

While Bullying/Harassment was generally down in the second application, the departmental breakdown analysis clearly shows where the organisation needs to focus its attention in addressing bullying behaviour, and minority groups are still raising issues in this area significantly more than their colleagues (p<0.05).

Dialogue

The frequency distributions of the items in this section appear normal as indicated by relevant descriptive statistics as well as boxplots which show symmetrical distributions and very few outliers. The degree of intercorrelation between these items is very good with a KMO value of 0.86, and the individual MSA values suggest that no item should be removed from this set. The PCA of the correlation matrix revealed only one general factor that accounted for 59% of the variance, and each item had a very high loading on this general factor. This suggests the item set is indeed a unidimensional scale. The analysis of the internal consistency of the items produced a Chronbach's alpha value of 0.86 suggesting that these items are homogeneous and can be added up in a sumscale.

This was the biggest success area for UKBankPLC, showing a dramatic increase in this score over the two time frames (from 42% to 69%). This suggests that the diversity awareness training has allowed people to discuss how they are feeling in the organisation, and engage in dialogue previously avoided. This is hugely important, as dialogue and communication can be the root to everything, as people talking generates understanding and allows people to express their difference and be accepted, rather than suppressing their difference in order to remain accepted.

Organisational Belonging

The descriptive statistics and boxplots show high means on all items, and for some items also peaked distributions, but otherwise there are no anomalies. The magnitude of the intercorrelations was high with a KMO value of 0.88. The PCA analysis revealed a very strong general factor which accounted for 70% of the total variance, and all items had a high loading on this general factor. This suggests these items form a strong scale, which was confirmed by the internal consistency analysis which revealed a Cronbach alpha value of 0.91 suggesting that this scale has excellent psychometric properties.

Not surprisingly, as dialogue increased, so did the sense of organisational belonging at UKBankPLC.

Emotional well-being (positive emotions)

The descriptive statistics and boxplots suggest normal distributions for these items and the degree of intercorrelations among these items is excellent with a KMO value of 0.91. Only one strong general factor emerged in the PCA analysis explaining 64% of the variance, and each item had a high loading on this factor. Hence this set of items can be regarded as unidimensional. An analysis of the internal consistency revealed a Cronbach's alpha value of 0.90 suggesting this scale also has excellent psychometric properties and can be used as sumscale with all the items being included.

Again, positive well-being also increased at UKBankPLC as would be expected as this is a positive outcome measure.

Emotional well-being (negative emotions)

Again the distributions of these items were symmetrical with very few outliers, and all relevant descriptive statistics fell into the normal range. The magnitude of the intercorellations was very good with a KMO value of 0.85 and the individual MSA values suggest that no items should be removed from the set. The PCA analysis extracted only one common factor which explained 59% of the variance and each item had a high loading on this general factor. Hence this set of items can be regarded as unidimensional and homogenous, as the analysis of internal consistency produced a Cronbach alpha of 0.88. No revision is therefore required to this sumscale and all items can be included.

Surprisingly, negative well-being remained almost constant at UKBankPLC which is counter-intuitive as it is a negative outcome factor. This could be explained, perhaps, by individuals being more aware of issues that could/should be affecting them and hence heightened sensitivity has resulted in increased distress. Alternatively, individuals could find the breaking down of their own stereotypes and challenging themselves and their personal prejudices stressful.

Discussion

In an effort to take a wider view of diversity climate and inclusion within organisations, the inclusion index has been developed to take account of 10 item sets and their impact on inclusion in the organisation. Senior managers set the tone of the culture and climate, outline the mission, strategy and vision, and influence the tone of voice that employees use to address each other. The immediate manager can have a direct impact in terms of day to day occurrences, handing conflict, communication and more general team management. The values of the individual and the organisation need to align to some extent, but equally do not have to be the same. Provided they are not conflicting then the individual should feel included in their workplace. To some extent, having a push towards 'shared values' can reduce diversity. If everyone shares the same values, then there is less room for difference. Recruitment policy and practice is a key area where an organisations approach to diversity becomes apparent, as it does with promotion, progression & development. These two areas are perhaps the most visible to the rest of the organisation as they can sees which 'groups' are getting the promotions, and which are being excluded. Hence these were the areas that UKBankPLC suffered lower scores with their second application of the index after diversity awareness training. Fitting in is true measure of inclusion - the extent to which an individual feels they 'fit in' to their workplace, and this is not a measure of conforming. Fitting in is about being accepted for who you are and fitting in regardless of any differences you may bring. Bullying and harassment are the visible extremes of discrimination action with an individual targeting and impacting negatively on another, often for a reason unknown to them. This is an area in which organisations hope to score lowly as efforts are widespread throughout organisations to deal with these areas, especially as they are widely covered by legislation. Dialogue is another measure of inclusion and relates to people having a voice that is heard, while organisational belonging relates to the embeddedness that follows fitting in. Finally emotional well-being takes both a positive and negative measure of how the workplace is affecting the well-being of the individual both mentally and spiritually.

The inclusion index is therefore a measure of both diversity and inclusion, and gives the organisation feedback on exactly where it needs to be targeting its efforts in order to improve both the performance of the organisation and the satisfaction of its employees. By running the index in an organisation before and after an initiative to develop inclusiveness, an organisation, such as UKBankPLC, can get a clear measure and evaluation of the impact of their initiative and how inclusiveness in building in their organisation, areas where there organisation needs to review its practice, and departments where issues are arising due to management practice that appears peculiar to the rest of the organisation.

Conclusions

This paper overviews the psychometric development and validation of a new multidimensional measure designed to assess diversity and inclusion within an organisation, and reports on its application over two time periods in UKBankPLC. The paper presents the results of the third iteration of the index as it has developed through a number of statistical trials prior to this publication. Overall, the findings of this study demonstrate acceptable levels of validity for the inclusion index to be used as an organisational measure of diversity and inclusion. However, as with any psychometric measure, development is by necessity an on-going process, and there are continuing needs for research into diversity and inclusion issues to identify new item scales in the future to further expand the index.