Research Design in Geography

With the ever-growing expansion of global knowledge geographers like many other scientists both human and physical have begin to face an “information explosion” (Ebdon, 1985). The readily available written information and numerical data today is increasing at an accelerating rate. This has lead to the necessity for summaries of these large data sets showing the concise measurements of their attributes. Human and physical geography can be seen as two different disciplines one been primarily focussed on qualitative data and one quantitative data respectively; both been equally reliant and interested within secondary data sources. The contention as to why this is the case is outlined below.

It can be noted that secondary data within the research discipline may be defined as “data which has not been collected with the specific research question in mind” (Emanuel and Egenvall, 2014). It can be seen as data which has previously been collected by somebody else however is effortlessly available; Secondary data was once a branch of primary data (Vartanian, 2010). It is seen as common source for academic disciplines to use within research projects either been obtained from quantitative or qualitative sources.

Secondary data as an aspect of scientific disciplines has come about due to the global widespread transition from paper to digital (Gomez and Jones, 2010). Nowadays datasets can be created, analysed and distributed worldwide digitally. Digital secondary data is often structured within databases and organised as tables which can be analysed. An example of this is weather records available online within the NOAA – National Climatic Data Centre (NCDC) whereby annual to daily summaries are available as structured datasets. This climatic data online provides free accessible archives of global historical weather and climatic data addressing all variables. Compared to many other scientific disciplines geographers use a great deal of secondary data because of the numerous types of data available within geographic research (Montello and Sutton, 2006).

Maps as a branch of secondary data support the basis of physical geographer’s research; there involvement in the environmental features, factors and processes which coalesce together to make a place unique. With the earth’s features and landscapes ever-changing in a spatial context maps allow geographer’s to study and monitor landscape change over time (Gabler et al, 2008). Climatologists are interested in weather maps as they show where and how weather elements change daily, over the seasons and yearly. This allows for predictions and management in areas which may be susceptible to high levels of rainfall, such as south West England; whereas Geomorphologists look at the study of the topography within a given area. Using maps physical geographers apply their knowledge they discover from the study of the earth; the observation of phenomena and compilation of data to seek solutions to the global issues to which they are interested in. Geographic Maps are readily available at (Digi-map) ordinance survey map data on Edina Maps; an example of a historic map can be found at

Physical geographers and other scientists work to describe and analyse the often complex features of planet earth and its environments by composing representations of the real world; models. A model is a simplified version of a more complex reality that allows for prediction; each model is designed with a specific purpose in mind. This is evident in Pacione (2001) with the idea of the concentric zone model and the multiple nuclei model in order to explain the spatial structure of the urban economy. A map is a branch of model as it shows a representation providing useful information required to meet specific needs. Maps are seen as a pictorial model (Gabler et al, 2008) and are used because they are efficient in conveying a great amount of spatial information that is easily recognisable. Likewise maps not only show spatial information and data but they also show essential information about the map itself which is interesting to geographers; the legend, scale and direction.

A more recent approach since the 1990’s is the use of geographic information systems and aerial remote sensing as a branch of mapping. Initially geographers used maps achieved by producing a transparent overlay for each data set at a common scale, aligning overlays so that their co-ordinates corresponded and then drawing a new overlay showing for instance how rock types and soils interrelate (Freeman et al, 1993). However the more data there is to analyse the more complex the map becomes as a piece of secondary data. Therefore large complex data sets require the use of computer software designed to manipulate spatial data. This computer based technology assists the large geographic data derived from numerous digital map layers (composed of thematic maps) enabling geographers to address global problems that require large amounts of spatial data from a variety of sources (Gabler et al, 2008).

Figure 1shows the clear procedure in order to use GIS to create secondary data.

As Moran (1975) states though it is not just geography using such tool, statistical analysis with maps is of course important in all other sciences such as geology, epidemiology and geophysics. An example of this is the work been done on the small scale geographical distribution of cases of tuberculosis, cancer and leukaemia (Moran, 1975). This is done to see if there is evidence of clustering; which may give light on the causation of such diseases.

Within the geographic discipline the common term secondary data refers to the relatively large databases those individual researchers would not be able to produce; for example census data, newspaper archives, satellite imagery or resource inventories. Secondary data is an important aspect in all geographic literature because it can be analysed in order demonstrate a depth of relationship between variables to show an underlying trend. Geographers use secondary data because it provides an alternative to the collection of primary data which in turn often gives the researcher access to more information than would be available (Vartanian, 2010). Figure 2 shows just a sample of large datasets available on Income Inequality as a branch of human geography.

Archives are seen as another branch of secondary data whereby the use of existing records that others have collected primarily for non-research purposes such as financial reports, birth and death records, newspaper stories, diaries or letters. These could be seen as more beneficial to human geographers as they are qualitative data.

A recent approach and use of secondary data is the use of personal solicited diaries as a qualitative method of research within social geography. In Meth (2003) diaries were used with women from South Africa who recorded their experiences of violence over a one month period. Within the article it shows that solicited diaries can contribute towards a feminist analysis of social processes similarly within human geography diaries can promote participation and engagement by respondents in the research process. This use of secondary data is also present in the recent study into the everyday geographies into the heterosexual love and home by Morrison (2012). Within this study solicited diaries are seen to provide participants with a sense of emotional reflection and they can allow researchers to access this knowledge which may not have been opened if another data collection method was used. However as Morrison (2012) states diaries offer “momentariness” research and cannot always be compared to everyday life.

Compared to many other scientific disciplines, both human and physical geographers use a great deal of secondary data (Montello & Sutton, 2006). Geographers can often be seen to study phenomena at large spatial and temporal scales where it can be seen as too difficult and upscale to collect data oneself. Likewise the idea that secondary data is not intended for ones research often inspires a geographers’ research area. As Montello & Sutton (2006) found, much geographic research is that analysts study problems at the examination scale of such available dataset, which is often not the scale at which the phenomena operates.

The primary reason for the use of secondary data is its availability; it is evident that there are thousands accessible in a myriad of places (Vartanian, 2010). This availability in such increasing amounts is due to the digitalization of many records. For human geographers the uttermost used source of secondary data is the population census (Flowerdew and Martin, 2013); which is produced in the UK every 10 years by the office of National Statistics (ONS). This in-depth data analysis provides demographic statistics but also details on education, transport, work and housing. Census data is available publicly online at no cost and is available globally; facilitating their use as an exploratory first step within a research project opposed to primary research within the same research area (Gomez and Jones, 2010). This approach is seen as more efficient in respect to time and cost in comparison to primary data collection. However it has become apparent that a large gap exists in the relative abilities of the rich and the poor countries to produce and control digital secondary datasets. However Gomez and Jones (2010) have seen the global south trying to narrow the digital divide by governmental projects in-order to create their own data collection. As geographers it is clear that the growing accessibility of digitalised data is related to the growth of geomatic technologies.

Emanuelson and Egenvall (2014) address the issue of time and cost; it is apparent that secondary data is cheaper and more readily available than primary data. Due to this the ability to gain large samples of data is seen more apparent likewise the chance to limit selection bias due to been able to sample a large part of the population. Primary data can be affected by specific biases such as recall and non-response. Secondary data is less likely to be affected by these biases due to the data been collected for another research question in mind. Questions should still be considered in secondary data such as how representative is the data, reliability and completeness of data to ensure validity. The data should be validated in the same way (i.e. Identification of non normal observations and internal validity). Reliability and validity are important questions within research as this offers consistency of results under repeatability conditions and offers a “truth-value of research” (Montello and Sutton, 2013).

The legitimacy of secondary data is carried by the organised order making it well suited for many types of quantitative or statistical analysis. Likewise secondary data is commonly produced by trained professionals who pre-test the questions and verify categories in order to produce standard and comparable information, both across time and space (Gomez, 2010). Most importantly the professional systems of collection assembly, storage and retrieval that constitute secondary data confer legitimacy that is widely recognised and works to empower such data and make it rhetorically convincing.

Secondary data can arguably be involved within geography due to the “Quantitative Revolution”; a term used by Davies (1972) as an aspect of one of the four major turning points within modern geography. This revolution occurred during the 1950’s and 1960’s highlighting a method of change behind geographical research; a launch from geography been a regional finding based research to a spatial science (Davies, 1972). The idea of secondary data been incorporated into the discipline meant that there was a movement from descriptive to scientific. As Davies (1972) states there is a still a divide between human and physical geography as it can be seen that physical has developed this “quantitative revolution” further causing a general talk of human geography becoming its own independent subject. The revolution itself is the basis for geography using secondary data today due to its creation of dynamism, self-insurance and a reassertion of scientific principles (Newby, 1980). This introduction of “scientific thinking” (Davies, 1972) engaged the geographic discipline into the solution for spatial, social and environmental global problems. By turning an introspective subject into an actively concerned discipline interested in the relationship it has with alternative global topics.

Throughout physical geography the growth in analysis has not only been linked to but also related to the change in content and focus of enquiry. An example of this is the growing use of systems and modelling approach in Geomorphology (Chorley, 1962) and the rapid expansion of technology allowing secondary data to be widely available. In human geography the beginning of quantitative techniques and the associated philosophical implications of a positivist approach led to change from 1965-75 (Gregory, 1983); arguably a decade later than physical geography. Urban geography experienced a drastic shift from an urban land use approach to quantitatively based studies of spatial urban and economic structures. This need for statistics within all aspects of geography was made clear by Wilson and Kirkby (1975) nevertheless some British Geographers are overlooking aspects which need real mathematical competence.

It has become clear that secondary datasets have become an important role in economic research due to the expansion of availability of datasets. Within human geography and economics international agencies such as the World Bank and the United Nations (UN) since the 1990’s have expanded its data sets, as for years have published income distribution data in its annual world development report. Advancements within these data sets are enabling a greater scale and distribution (Atkinson and Brandolini, 2001). An example of this can be seen by the data sets constructed by Klaus Deininger and Lyn Squire (1996) and the world income inequality database (WIID).

Alongside the expansion of research it can be noted that research has changed over time. This is displayed in the Social Service Review (SSR) during 1980 and 2007. In 1980 six main articles or notes used some form of secondary data either administrative or survey data whereas in 2007 it was twenty-two used articles published. Vartanian (2010 argues this is only a snapshot of a trend based on one elite social work journal; However secondary data is becoming increasingly important.

Statistical data is an important aspect of geography as it offers credibility to an argument or advice. Moran (1975) claims that statistical geography bears the same relation to geography that econometrics do to economics. Statistics are present in all academic journals and are constantly been generated by governmental organisations in-order to generate spatial trends. Governmental run datasets such as national statistics online, the UK population census and GEsource, all offer data which can be found across most countries and can usually be disaggregated to quite small areas such as administrative and political divisions; which are popular amongst geographers (Flowerdew and Martin, 2013). Moran (1975) discovers that a great deal of statistical geography appears to be more descriptive than explanatory.

The most common use of statistics in the UK by human geographers is the population census data. A geographic use of this is using census data to look at migration and morbidity in Bentham, G, (1988). Census data is an official complete collection of data from the population with details as to age, sex and occupation and renewed every 10 years. Bentham, G (1988) looks at the association between the geographical pattern of disease and possible casual factors; looking at the 1981 GB Census data. Self-reported morbidity statistics are used; displaying that the health status of migrants differs noticeably from that of non-migrants. Similarly Mesev, V (1998) uses census data within urban image classification. Mesev looks at a monitored classification strategy containing a group of techniques that allow the connecting of urban land cover from remotely sensed data with urban functional characteristics from the population census data.

However statistical data should not bind us solely to secondary data; in addition there are administrative reports, business records, diaries, newspapers and maps. As with any form of methodology disadvantages are there; secondary data can be argued to have a lack of control (Vartanian, 2010). It can be said to have a lack of control over the framing and wording of survey items and that the questions important to your studied may not be included in such data. Likewise subtleties matter a great deal in research and secondary data can be argued to get broader and not answer the research question in the direct research title. Similarly Emanuelson and Egenvall (2014) consider that there is no control over the information what is included in datasets which have already been produced therefore impossible to validate. Moreover ecological fallacy and modifiable area unit problem can be an issue within secondary data; the assumption that all individuals in a group share the average characteristics of that group and those trends within data are based upon existing boundaries that are unrelated to the phenomena in question.

Secondary data will remain important to geographic research as a primary source of information to a growing number of data intensive applications. Using secondary data clearly gives the researchers important advantages such as data coverage, quality and costs as well as the ability to analyse phenomena that otherwise may be impossible such as analysis of populations at a global scale. It can be argued that “Data” refers to a body of information in numerical form therefore it can be argued that it is hard to categorise data as uniquely geographical except perhaps data which concerns the spatial characteristics of places and areas (Ebdon, 1985). GIS as a branch of mapping is seen as one of the basic uses of secondary data within physical geography due to its ability to provide an important route to enquiry enabling exploration and integration of geographical data (Freeman et al, 1993).

Within the immediate future physical geographers have no sign of movement away from the statistical analysis and the importance of using secondary data and mathematical modelling is more likely to grow opposed to contract. Whereas within human geography the future is less clear; the positivist view point is being challenged leading to a number of coexisting approaches. Nonetheless according the Institute of British Geographers for the future “the numbers game is far from over” (Newby, 1980) and this analogy can be applied to many scientific disciplines.