Monthly Archives: October 2014

Interim Survey Data Analysis: Conclusion

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

 

The preceding analysis is interim in nature and leaves much room for improvement, not least in terms of specification of the variables included in the models. Still, if the variables used are accepted as rough indicators of the concepts that are at play then some initial observations can be made. First amongst these is that the Civic Voluntarism Model does a good job of accounting for the frequency of political activity, and performs better in this regard than the Bourdieusian and perceptual model. Those components do add a small amount of power to the Civic Voluntarism Model, when included in the full model, but they leave a lot to be desired from the perspective of parsimony.

Much of the power of the Civic Voluntarism Model stems from the inclusion of the group recruitment variable, which is by far the most powerful predictor of political activity in the full model. Questions remain around the proximity of the question underpinning this variable to the dependent variable in the original survey but, assuming that it is not compromised, it is clear that Verba, Schlozman, and Brady were right to identify it as a component of their model. In fact, the strength of its relationship with the dependent variable suggests that it may be worth running separate analyses in which recruitment is treated as the dependent variable. Further, if recruitment is such a strong predictor, it may be worth investigating whether there are other different processes at play across groups that are frequently or infrequently recruited, for instance by comparing separate analyses across those groups.

If the performance of the Civic Voluntarism Model, and recruitment in particular, is difficult to ignore this should not lead us to overlook the consistently active variables in all four elements of the full model. These robustly significant relationships indicate further interesting results that may need to be teased out through more complex analyses. Such analyses should focus first on specifying improved variables, based on the available survey questions, through confirmatory factor analysis. Such variables could then be utilised in structural equation modelling that has the capacity to accommodate a more complex structure of relationships between variables than does multiple regression. These are the next steps for the current research.

Until those analyses are undertaken, however, we can take away some interim observations about particular variables from the current analysis. First, identifying as non-heterosexual, considering oneself to be limited by a disability, and having health conditions are all associated with greater political activity. Second, exercising key civic skills at work, being interested in politics and discussing it, and being asked to get involved in voluntary groups are all positively related to political activity. Further, interest in politics and group recruitment are two of the strongest predictors of political activity. Third, having an array of cultural tastes is negatively associated with political activity but being culturally active is positively associated. By contrast, having a range of social acquaintances is positively associated with political activity. Fourth and finally, perceiving the role of privilege in structuring society is positively related to political activity. Thus, whilst Verba, Schlozman, and Brady’s Civic Voluntarism Model does an impressive job of explaining political activity, it seems that parts of the Bourdieusian and perceptual model also have a part to play.

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

Back to Academic page

 

Interim Survey Data Analysis: Hypotheses and Analysis of Models

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

 

As the overarching title of this post suggests, the analysis presented here should be treated as interim. Many of the variables outlined above are approximations of concepts and require considerable refinement, particularly in terms of subjecting their components to confirmatory factor analyses. The use of multiple regression is also a starting point, with the aim being to move onto structural equation modelling in the hope that it can accommodate the complexity of the theoretical models. Still, the results provide some food for thought in the sense that they identify variables in all four elements of the model that are robustly important in accounting for levels of political activity.

The main purpose of this analysis, before identifying consistently significant independent variables, is to consider how well the four elements of the full model account for political activity. This is achieved, first, by cumulatively adding those elements to the model and observing their impact on the Adjusted R-Square (more conservative than its non-adjusted counterpart), or proportion of the variation in the dependent variable accounted for by the independent variables. As such the hypotheses that are being tested in the interim analyses are as follows:

 

Hypothesis 1 – Adding the variables representing the forms of capital will improve the Civic Voluntarism Model in terms of the amount of variation in the dependent variable accounted for;

 

Hypothesis 2 – Adding the variables representing perception of privilege will improve the Civic Voluntarism Model in terms of the amount of variation in the dependent variable accounted for;

 

Hypothesis 3 – The Bourdieusian capital and perception of privilege model of political participation will perform as well as the Civic Voluntarism Model in terms of the amount of variation in the dependent variable accounted for.

 

The results of the cumulative analyses are shown in Table 9, which presents the variables relating to background characteristics as the baseline model. The components of the Civic Voluntarism Model were then added in the second model, followed by the additional Bourdieusian economic, cultural, and social capital variables in the third model. Finally, the variables relating to perception of privilege were added to create the full model.

Despite including variables that are statistically significant (highlighted in bold and italics) at the 10% level or above in terms of their impact on political activity, the baseline model (Model 1) does not perform well in terms of Adjusted R Square (0.083) and accounts for less than 10% of the variation in the dependent variable. Things get more interesting when the variables from the Civic Voluntarism Model are added (Model 2). Again, some of the variables are significant at the 10% level or above but the real action is in the impact on Adjusted R Square, which shoots up to 0.460. This means that, statistically speaking, the components of the Civic Voluntarism Model account for a full 37% more of the variation in political activity than do background characteristics alone.

Less impressive is the effect of adding the Bourdieusian elements of economic, cultural, and social capital to the model (Model 3). Some of those variables are significant at the 10% level but, combined, they only increase the Adjusted R Square to 0.486. The 13 variables in the Bourdieusian element of the model, then, only account for 2.6% more of the variation in the dependent variable than do background characteristics and the Civic Voluntarism Model alone. Finally, the addition of the three variables relating to perception of privilege (Model 4) raises the Adjusted R Square only a tiny amount (to 0.489). These results are summarised in Chart 3, in which each iteration of the model includes an indication of the number of variables in it (in parentheses), and the full model is labelled as ‘Civic Voluntarism, Bourdieusian, and Perception’.

 

Table 9 – Cumulative Addition of Elements to Construct

Full Model Accounting for Political Activity

(please click on the table for a full-size version)

2015-08-13 Corrected Interim Survey Data Analysis Table 09

 Chart 3 – Adjusted R Squares of the Cumulative Models

2014-10-14 Interim Survey Data Analysis Chart 03

 

The fifth model in Table 9, labelled as ‘Parsimonious Full’ in Chart 3, includes only the significant (at the 10% level) variables from the full model. As can be seen this model includes fewer variables than all of the others except the baseline model but it performs just as well as the full model in terms of Adjusted R Square (0.493). This means that with just 13 variables we can account for over 49% of the variation in the dependent variable. Still, the parsimonious model performs only slightly better than the Civic Voluntarism Model, which seems to account for most of the impressively high Adjusted R Square. However, from the cumulative models we cannot observe how well the elements of the full model perform on their own and, therefore, in comparison to each other.

Of particular interest when treating the elements of the full model as separate models in their own right is the performance of the Civic Voluntarism Model compared to that of the Bourdieusian and perceptual model. This is because the former can be considered the established model whereas the latter is a new one drawing on the theories outlined in the literature review. Each model is tested in various forms but the variables from the baseline model are included in all instances. This is because they are widely used control variables and because they represent important background characteristics in their own right. The results of these comparisons appear in Table 10.

The focus again is on amount of variation in the dependent variable, which remains the frequency of political activity, that is accounted for by the independent variables in each model. The first model (Model 2a) is the Civic Voluntarism Model with the group recruitment variable removed. This has been done because the power of that variable (Beta = 0.445) may indicate problems with it. In particular, the question that it is based on is only separated from the question that underpins the dependent variable by three other questions. Further, those three questions are themselves about levels of political or voluntary activity, and requests to engage in such behaviour. As such there is a risk that respondents, having just answered questions on their levels of political and voluntary activity, will have subconsciously altered their answers to proximate questions on recruitment so that they reflected those levels of activity more closely. The power of the group recruitment variable can be seen in the reduction of Adjusted R Square (from 0.460 to 0.301) when it is removed from the Civic Voluntarism Model (compare Model 2 with Model 2a), and it is for this reason that the subsequent models are tested both with and without it included.

The parsimonious Civic Voluntarism Model (Model 2b), including the significant variables from the full equivalent, is tested next. Given that it only includes eight variables it has a remarkably high Adjusted R Square (0.463), though a large proportion of that comes from the inclusion of the group recruitment variable. Still, this is by far the best performing model if the number of variable included is considered alongside the Adjusted R Square. By contrast, the model including the variables representing Bourdieu’s economic, cultural, and social capital as well as the perception of privilege variables (Model 4b) accounts for 20% less of the variation in the dependent variable (Adjusted R Square = 0.264) than does the parsimonious Civic Voluntarism Model. Even if the recruitment variable is included alongside the Bourdieusian and perception variables (Model 4c) they do not perform as well as the Civic Voluntarism Model (Adjusted R Square = 0.434). The parsimonious versions of the Bourdieusian and perceptual models excluding recruitment (Model 4d) and including it (Model 4e) perform only marginally better (Adjusted R Squares of 0.273 and 0. 438 respectively).

 

 Table 10 – Comparison of Elements of the Full Model

as Separate Models Accounting for Political Activity

(please click on the table for a full-size version)

2014-10-14 Interim Survey Data Analysis Table 10

 

The results of the model comparison are summarised in Chart 4, which again indicates the number of independent variables in each model (in parentheses), and includes the full Civic Voluntarism Model (from Table 9) for reference. The Bourdieusian and perception models perform consistently worse than the various iterations of the Civic Voluntarism Model. If a conservative approach is taken and the group volunteering variable is excluded (because of the question’s proximity, in the survey, to the question underpinning the dependent variable) then the gap is narrowed to a very great degree. Indeed, it accounts for between 2.8% and 3.5% less variation (respectively, in its parsimonious (Model 4d) and non-parsimonious (Model 4b) forms) in the dependent variable than does the Civic Voluntarism Model excluding the group recruitment variable. This gap is sustained if the variable is included in the parsimonious versions of both (Model 2b and Model 4e). However, if a strict reading of the texts that inform models is adhered to then the group recruitment variable should only be included in the Civic Voluntarism Model, making it much stronger than its Bourdieusian and perceptual counterpart.

 

 Chart 4 – Adjusted R Squares of the Comparative Models

2014-10-14 Interim Survey Data Analysis Chart 04

 

In terms of the three hypotheses, the first two are supported to a very small degree whilst there is no evidence emerging to support the third. That is to say that adding the Bourdieusian and perception of privilege variables do improve the Civic Voluntarism Model, marginally, in terms of the amount of variation in the dependent variable that is accounted for. However, the number of variables added to gain such a small increase in Adjusted R Square suggests that the Civic Voluntarism Model alone is preferable. This is further supported by the fact that none of the analyses indicate that the Bourdieusian and perceptual model alone performs better than the Civic Voluntarism Model.

Although the preceding seems pretty conclusive there are hints of other paths that may be worth following. The fact that the Bourdieusian and perception of privilege models perform only slightly worse than the Civic Voluntarism Model suggests that they may be variables of interest included in the former, which is confirmed if Table 9 is reviewed in more detail. This reveals that four of the variables in the baseline (background characteristics) model remain significant at the 10% level or above regardless of the other variables that are added to the model. Taking the results of the parsimonious full model are it appears, first, that respondents who did not identify as heterosexual were more likely to be politically active than those who identified as heterosexual (Beta for the heterosexual binary variable was -0.067). Although the Bs are difficult to interpret because the dependent variable is the sum of eleven different political activities, it can still be observed that heterosexual respondents had an average score that was over two points lower than their counterparts who did not identify as heterosexual (B = -2.276).

Still focusing on the parsimonious full model, having some level of limitation due to a disability was also related to a greater level of political participation (Betas for the binary variables indicating a little limitation and a lot of limitation were, respectively, 0.051 and 0.065). Respondents with a little limitation, on average, had a dependent variable score that was more than one point higher than respondents with no limitation stemming from a disability (B = 1.130) whilst those respondents with a lot of limitation had a score that was a approaching two points higher (B = 1.849). Similarly, having a larger number of health conditions was positively related to political activity (Beta = -0.065). The negative sign preceding the numbers relating to this variable appear because it has been inverted so that a higher score equates to being healthier (i.e. having less health conditions). Thus, amongst the respondents to the survey, having an additional health condition was related to an increase of almost three quarters of a point on the dependent variable scale (B = -0.665, again the negative sign is present because the independent variable is inverted). These findings may appear counterintuitive and any interpretation at this stage is speculative, however all of the groups that have higher levels of political activity either have a history of being socially marginalised, or have practical issues in their day-to-day lives that may prompt them to be more active in pursuing influence.

Moving on to the active components of the Civic Voluntarism Model we can see that work based civic skills are positively related to political activity, though the effect is small (Beta = 0.062) and a one point rise on the civic skill scales is associated with a rise of less than a tenth of a point (B = 0.080) on the dependent variable. By contrast, political interest has a much stronger positive relationship with political activity (Beta = 0.272) and, in fact, is the second strongest predictor in the parsimonious full model. On average a one point rise on the combined political interest scale is associated with approaching a half point (B = 0.457) rise in the rough measure of political activity used here. Only group recruitment has a stronger positive relationship with political activity (Beta = 0.411) than political interest. In fact, as mentioned, group recruitment is by far the strongest predictor of political activity, and a one point rise on the group recruitment scale is associated with over a half a point rise in the dependent variable (B = 0.547). The three active components of the Civic Voluntarism Model, then, all have the anticipated positive relationships with the dependent variable, and political interest and group recruitment are the two variables in the parsimonious full model with the strongest relationships with political activity.

As can be discerned from the comparisons of the various models above, the active components of the Bourdieusian model are less powerful than those in the Civic Voluntarism Model. Nevertheless, there are some interesting relationships at play. The count of cultural tastes variable has a negative relationship with political activity (Beta = -0.116), which is to say that having more encompassing cultural tastes is associated with being less politically active. This is not an especially strong relationship as it is a non-integer variable running from zero to three, so a full one point change on the scale is a great deal of movement. Still, such a change is associated with something approaching a one and a half point lower score on the dependent variable (B = -1.399). Slightly stronger is the impact of the frequency of cultural activities (Beta = 0.161), which is positively related to political activity. This is a summary variable with a long scale and a one point change on it is associated with only a very small change on the dependent variable (B = 0.077). Together, the impact of these measures of cultural capital suggests that those people who are more frequently culturally active are, perhaps unsurprisingly, also likely to be more politically active. However, the idea that cultural omnivores, with wide ranging tastes, are also more likely to get involved in politics is not supported by this analysis of the survey data.

Moving on to social capital, the count of the number of categories of people, based on job status, that respondents know is positively associated with political activity (Beta = 0.077). A one-point change on this variable, which is to say being acquainted with someone from one more category of job status, is associated with around a fifth of a point rise in political activity (B = 0.181). Lastly in the Bourdieusian components of the parsimonious full model, the diversity of respondents’ acquaintances is positively related to political activity (Beta = -0.066, again the variable is inverted so a negative sign indicates a positive relationship). This variable has a short constructed non-integer scale meaning that a full one point change indicates a large difference. Such a change is associated with a two-thirds of a point change on the scale indicating political activity (B = -0.693). Both of these relationships support the idea that being acquainted with more groups of people increases political activity. Crucially, there is no evidence emerging from the current analysis to suggest that knowing people with particular statuses is important, rather it seems to be having a wide set of acquaintances that plays a role.

Finally, and reassuringly in light of the focus of the research, higher perception of the role of privilege in structuring social hierarchies is associated with greater political activity (Beta = 0.060), though not especially strongly. A one point rise on the measure of perception of the importance of privilege in society is associated with approximately a quarter of a point rise in the dependent variable (B = 0.242). This is an interesting finding and suggests that those who see a more important role for privilege in structuring society are apt to undertake more political activity. Again, interpretation at this stage is speculative but the results suggest that perceiving privilege in society motivates a desire to influence politics. Whether this is tendency is stronger amongst those with higher or lower levels of privilege is a question for further analysis but those possibilities suggest quite different motivations.

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

Back to Academic page

 

Interim Survey Data Analysis: Survey Items

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

 

The first wave of the survey contained 72 questions, and the second wave contained 100 questions. These were distributed between the three sections in each wave as shown in Table 8, and many of them were filtered questions. This means that they were only asked if particular answers to previous questions had been chosen by respondents. In some cases this meant that no question was asked and in others it meant that only one version of multiple alternative questions was asked. Further, some of the filtered questions were only asked of respondents if their answers had not already been recorded in previous surveys. As such, it would be time consuming to establish how many questions respondents received on average but it is the case that no respondent would have received all 172. It is also the case that the 1501 respondents in the sample provided answers to all of the questions that they received because progression through the survey was conditional on providing valid answers. This means that the combined first and second wave answers provide a rich picture of the respondents.

 

Table 8 – Number of Questions and Filtered Questions

in the Sections of Each Wave of the Survey

2014-10-14 Interim Survey Data Analysis Table 08

 

As outlined in the introduction there are four elements of the full model analysed below. These are background characteristics, the components of the Civic Voluntarism Model, further elements of economic, cultural, and social capital, and perceptions of privilege. Before the analysis of those elements is presented it is useful to outline the variables that constitute each of them. As will become clear, some of the variables are simply single survey items whilst others are based on multiple items. A number of these variables are likely to be respecified in the course of further analysis. As such, the below variables reflect the interim nature of this analysis.

 

Background Characteristics:

Age (years) – As the name suggests, this is a simple interval variable indicating respondent age in years.

Female (binary) – Again, as the name suggests, this is a binary recoding of the sex variable with ‘Female’ coded as 1 and ‘Male’ as 0.

White British ethnicity (binary) – This is a recoded version of the ethnicity variable included in the data with ‘White British’ coded as 1 and all other ethnic groups as 0. The latter includes the ‘White Other’ group because it is considered that such respondents are likely to be European migrants who, as such, should be considered distinct from the majority population.

Religious (binary) – This is a recoded version of the religious denomination variable included in the data with all of the religious options combined into the 1 category and the remaining non-religious categories combined into the 0 category.

Heterosexual (binary) – This is a recoded version of the sexuality variable included in the data with ‘Heterosexual’ recoded as 1 and all other sexualities as 0. The ‘Prefer not to say’ answers were recoded into 9 and excluded from the analysis.

Disability limitations (three categories) – This is an inverted version of the original three-category variable in the data so that 2 represents significant limitations resulting from disability, 1 equates to some limitations, and 0 indicates no limitations. The categories have been split into binary variables in the model with the no disability limitation category acting as the reference.

Health Conditions (count) – This is a summary variable inverting the count of positive (with a value of 1) selections of each category of health condition, meaning that a higher value in this (with a maximum possible of 19) represents having fewer of the listed health conditions.

 

Components of the Civic Voluntarism Model:

Educational qualification (highest) – This is an ordinal variable included in the data, with higher numerical values (there are 18 categories) associated with higher educational qualifications. The categories have been split into binary variables (with the ‘No qualifications’ acting as the reference category) and the ‘Don’t know’ and ‘Prefer not to say’ answers are excluded from the analysis.

Gross household income (fifteen categories) – This is an ordinal variable included in the data, with higher values (as the name suggests, there are 15 categories) indicating a higher income bracket. The categories have been recoded into binaries for the analysis (with the lowest income acting as the reference category).

Free time (hours per weekday) – This is an interval variable included in the data indicating the average number of free hours per weekday reported by the respondent (after sleep and a range of responsibilities such as work and childcare are accounted for).

Work-based civic skills – This is a calculated summative variable indicating the frequency of undertaking four key skills in current or past work. The highest possible value (28) indicates that all four of the skills (writing a formal email or letter, taking part in a decision-making meeting, chairing or planning a meeting, and giving a presentation) are used every day. The original answers were inverted before summing so that the ‘Daily’ answer had the highest value.

Political interest – This is a calculated summative variable indicating the level of attention paid to local and national politics, and the frequency with which each is discussed. The component variables were inverted so that high interest and frequent discussion had the highest values before they were summed. This means that the highest possible value in the summative variable (24) indicates ‘A great deal’ of interest in both local and national politics, and that both topics are discussed ‘Every day or almost every day’.

Political efficacy (internal and external) – This is a calculated summative variable indicating both the belief that citizens have influence over the political system, and the respondent’s self-assessed level of influence relative to others, as well as their ability to understand politics. The latter variable was inverted before summing so the maximum score (36) on the summative variable indicates that respondent strongly disagrees that it is often difficult to understand what is going on in government and politics, that they believe members of the public have ‘A great deal’ of influence at local, regional, and national level, and that they believe they have ‘Much more’ influence than most people at all three of those levels.

Political knowledge – This is a calculated summative variable indicating locally relevant political knowledge and national government knowledge. The variable is coded so that the maximum value (9) indicates correct identification of the local MP (from a list), correct answering of a question relating to whether MPs have to answer letters, correct identification of five senior politicians (of varying renown), and certainty of identifying a local group meeting space.

Identifies with a party (binary) – This is a recoding of the nominal party identification variable included in the data so that identification with any party is assigned a value of 1 whilst lack of identification is coded as 0.

Group recruitment – This is a calculated summative variable indicating the frequency with which invitations to get involved in voluntary groups are received via a range of means. Each of the constituent variables was inverted before summing, meaning that the maximum possible score (44) indicates requests being received ‘Once a month or more often’ from each of eleven possible sources (mass communication, family member, friend, neighbour, colleague, member of a religious congregation, political party member (if respondent is a member of a party), trade union or professional association member (again, if respondent is a member themselves), campaigning organisation member (if respondent is a member), charity member (if respondent is a member), or campaigner from an organisation that the respondent is not a member of).

 

Bourdieusian Elements of Economic, Cultural, and Social Capital:

Count of cultural activities – This is a calculated summative variable totalling coded binary variables indicating whether a number of cultural activities are undertaken outside the home, in the home, and when on holiday. The variable is itself a sum of three summative variables relating to each of those areas. Because the three constituent scales are of different lengths (there are more external activities to choose from than there are home-based activities or holiday activities) they are divided by the maximum score possible in each case so that none of the three scales can contribute more than 1 to the overall summative scale, which therefore has a (non-integer) numerical value of between 0 and 3.

Count of cultural tastes – This is a calculated summative variable totalling a sequence of binary variables indicating tastes in different areas. These are the venues visited when eating out, the cuisines consumed when eating out, the radio stations listened to, the types of magazines read, the television channels watched, the music genres selected, and the film genres selected. As with the previous variable, the overall summative variable in this case is a sum of five constituent summative variables, each of which has been divided by the maximum score possible in each case so that they contribute equally. As such, the overall summative variable has a (non-integer) score of between 0 and 5.

Frequency of cultural activities – This is a calculated summative variable totalling the frequency of a range of cultural activities taking place outside the home (from playing bingo to attending the opera), inside the home, and of holidays. The constituent variables were inverted so that frequent attendance was coded with a higher score. This is a simple overall frequency variable so greater frequency of any cultural activity will contribute to a higher overall score. As such, the maximum score (160) on the summative scale indicates that all seventeen external activities are done ‘A couple of times a month or more often’, all eight home-based activities are done ‘Daily’, and holidays are taken ‘Twice a year or more often’ with all four possible groups of people.

Legitimate status of activities – This is a calculated summative average variable. The frequency of each of the external, internal, and holiday cultural activity was multiplied by a score of between 1 and 3 (e.g. bingo and opera, respectively, in the case of external activities), those scores were summed and then divided by the count of cultural activities participated in to give an average legitimate status of activities taking into account their frequency.

Legitimate status of tastes – As above, this is a calculated summative average variable. The eat out venues, radio stations listened to, magazine types read, television stations watched, music genres selected, and film genres selected were assigned a score of between 1 and 4 (e.g. fast food outlet and fine dining restaurant, respectively, in the case of eating out venue), those scores were summed and divided by the count of those selected to give an average legitimate status of tastes.

Friends and acquaintances – This is a calculated summative variable totalling the numbers of friends seen daily, weekly, and monthly with the number of close relatives living nearby and further away, and the number of neighbours known. The maximum score (24) indicates that the highest option was opted for in each case, with no inversion of the original variables needed to achieve this.

Intensity of Relationships – This is a calculated summative variable totalling the frequency of seeing friends and acquaintances as well as the level of support received from various groups. The option selected for number of friends seen daily was multiplied by three (allowing a score of up to 12), for number of friends seen weekly by two (allowing a score of up to 8), and monthly by one (meaning a maximum score of 4). To these variables was added the inverted frequency of talking to neighbours (maximum score of 4), the inverted frequency of going out with colleagues (maximum score of 7), and the number of thirteen different types of help received from five groups (giving a maximum score of 65). This meant that of the total maximum summative score (100) considerably more than half could come from the help received by respondents. This was an intentional weighting in the summative variable towards the measurement of what can be relied on in relationships rather than just the frequency of engaging with acquaintances.

Count of acquaintances status categories – This is a simple calculated summative variable totalling the number of binary selections indicating acquaintance with people holding a range of job statuses. The maximum possible score is 14.

Diversity of acquaintances – This is a calculated summative variable reflecting the proportion of friends from the same gender, ethnicity, and religion as the respondent. In the case of gender, fifty was set as the most diverse point on the percentage scale, meaning that a new variable was calculated with fifty set as zero (with zero becoming minus fifty and one hundred becoming fifty) so that the numbers could be squared to give a curve with its lowest point at zero (reset from fifty). Thus, a score higher than zero indicated that more of one gender was known than of the other. This score was then divided by two thousand to give a maximum possible value of 1.25, which was then added to the proportion of friends from the same ethnicity and from the same religion, each divided by eighty (again giving a maximum possible value of 1.25 in each case). The variables relating to ethnicity and religion, each of which (unlike gender) has notably more than two possible groups, were not recoded (except for being divided by eighty) because it is considered that a lower proportion of friends from the same ethnicity or religion increases the likelihood of a diverse friendship group. This is a flawed assumption (the diversity is likely to depend on whether the respondent is in the majority group in society) but is considered acceptable for the interim analysis. As such, the maximum possible score on the summative variable is 3.75. This value was opted for as the maximum so that, if combined with the previous variable it would contribute less of the overall variation (because of the flawed assumptions involved in calculating this variable).

Average status of acquaintances – The fourteen different categories of acquaintances’ jobs were recombined as necessary to give the nine categories reflecting the NS-SEC and then assigned a weight on the basis of their status. The weighted values were then summed and divided by the count of categories selected by respondents to give an average of the status of acquaintances, with the maximum possible score being 8.

Property ownership – The five binary variables indicating ownership of a range of forms of property were weighted between 1 and 3 (depending on the asset type (e.g. ownership of commercial premises was weighted with a 3 whereas holding a mortgage on a property to live in was weighted with a 1)) and the scores summed to give this variable with a maximum score of 10.

Assets (fourteen categories) – This is an ordinal variable included in the data, with higher values (as the name suggests, there are 14 categories) indicating a higher income bracket. The categories were recoded into binaries for the analysis (with the lowest asset value acting as the reference category).

State benefits received – This is a simple constructed summative variable totalling the binary variables indicating receipt of a range of state benefits. There are twelve binary variables so the summary variable has a maximum possible score of 12 (indicating receipt of all listed state benefits).

 

Perception of Privilege:

Perceived role of privilege in society – This is a calculated summative variable based on the binary selection and inverted rank score of privilege-orientated explanations of status differences in society. One point was assigned if either ‘Because of their backgrounds’ or ‘ Because of inequality based on things like sex, race, and religion’ was selected from a list of possible explanations for status differences in society. A further three points were assigned if either was subsequently ranked as the top explanation, two points if ranked second, or one point if ranked third. This allows a maximum possible score on the summative variable of 7 (both privilege-orientated explanation selected, one ranked first, and the other second).

Perceived role of privilege in own life – This is a summative variable that was calculated in a similar manner to the previous variable. However, because there was no initial multiple-choice question in this case the summary score is based only on the ranking of the background or structural inequality explanations as they apply to the respondent’s life. In addition, a subsequent filter question went to those who did not rank either option asking them whether they thought background played any part in their status. If they answered yes to this prompt then they were assigned half a point. As such, the maximum possible score on the summative variable is 5 (ranking background first and structural inequality second, or vice versa, as explanations for own status, in which case the subsequent prompt question would be skipped).

Perceived status in society and own group – This is a simple calculated summative variable totalling the respondent’s self-placement on two ten-point hierarchical scales (respectively relating to their status in society at large and within the group of people they know). The two component variables were inverted so that high values reflect high self-perceived status, and the summative variable has a maximum score of 20.

 

Dependent Variable:

Frequency of political acts – This is a summary variable of the frequency of undertaking eleven different political acts (from displaying materials to going on a protest). The original answer categories have been inverted so that higher frequency has a higher value. As such, the maximum value of 55 indicates that all eleven activities are done ‘Once a month or more often’.

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

Back to Academic page

Interim Survey Data Analysis: Sample Information

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

 

The YouGov panel is comprised of more than 360,000 respondents who have opted-in to answer surveys for the company. Such respondents are recruited through the YouGov website as well as through advertising on other websites, and the company makes extensive efforts to recruit underrepresented groups to the panel. Nevertheless the sample, though drawn at random from the YouGov panel, is not a random sample of the United Kingdom population. Comparing the sample to the 2011 Census data we can see that the sample is broadly representative of the British population in terms of sex and region of residence. However, the sample is less representative in terms of age, ethnicity, and religion.

YouGov does not recruit panellists who are below the age of 18 which explains the low proportion of respondents in the 15-19 cage category defined by the Census. Putting this aside it remains the case that younger people (below the age of 30), older people (over the age of 75) and, less dramatically, people in their forties are underrepresented. By contrast respondents in their thirties and especially in their fifties and sixties are overrepresented in the sample. In terms of ethnicity, White British people are overrepresented in the sample whilst almost every other ethnic group is underrepresented. The story is a similar one in relation to religion, with respondents with no religion overrepresented and all other religious groups underrepresented in the sample.

 

Table 3 – Comparison of

Sample and Population Sex

2014-10-14 Interim Survey Data Analysis Table 03

 

           Table 4 – Comparison of Sample                                Table 5 – Comparison of Sample

       and Population Region of Residence                            and Population Age Categories

2014-10-14 Interim Survey Data Analysis Tables 04 and 05

           Table 6 – Comparison of Sample                                Table 7 – Comparison of Sample

              and Population Ethnic Group                                   and Population Religious Group

2014-10-14 Interim Survey Data Analysis Tables 06 and 07

 

The fact that the sample is not representative of the British population may have implications for any findings emerging from the subsequent analysis, especially if the goal is to generalise those findings to the population. However, representativeness of samples is more important when presenting descriptive statistics than it is when examining causal relationships. This has been demonstrated in an article by David Sanders, Harold D. Clarke, Marianne C. Stewart, and Paul Whiteley.[1] They conducted a survey experiment in which a face-to-face survey was fielded to a clustered probability sample whilst a self-completion survey containing the same questions was fielded online to a sample of YouGov panellists. Analysis of the resulting data revealed significant but minimal differences in the levels of party support indicated by the two samples. By contrast, the significant variables emerging in models of voting behaviour were almost all the same. Based on those findings, and given that the goal of the current research is to test models of political behaviour rather than to describe distributions in the population, it is reasonable to utilise the survey data gathered despite its lack of representativeness.

Sections: Introduction -> Survey Details -> Sample Information -> Survey Items -> Hypotheses and Analyses of Models -> Conclusion

Back to Academic page

 


 

[1][1] David Sanders, Harold D. Clarke, Marianne C. Stewart, and Paul Whiteley, ‘Does Mode Matter for Modeling Political Choice? Evidence from the 2005 British Election Study’, Political Analysis, Vol. 15, No. 3 (Summer, 2007), pp. 257-285.