Publication - Research and analysis

Exploring Dimensions of Social Capital in Scotland Findings from the Scottish Social Attitudes Survey and Scottish Household Survey

Published: 29 Oct 2012
ISBN:
9781782560951

The report explores whether different groups in society experience different levels of social capital. It draws on data from the Scottish Social Attitudes survey (SSA) 2009 and the Scottish Household Survey 2010.

Exploring Dimensions of Social Capital in Scotland Findings from the Scottish Social Attitudes Survey and Scottish Household Survey
Annex A - Regression models

Annex A - Regression models

1. Regression analysis aims to summarise the relationship between a 'dependent' variable and one or more 'independent' explanatory variables. It shows how well we can estimate a respondent's score on the dependent variable from knowledge of their scores on the independent variables. This technique takes into account relationships between the different independent variables (for example, between education and income, or social class and housing tenure). Regression is often undertaken to support a claim that the phenomena measured by the independent variables cause the phenomenon measured by the dependent variable. However, the causal ordering, if any, between the variables cannot be verified or falsified by the technique. Causality can only be inferred through special experimental designs or through assumptions made by the analyst.

2. All regression analysis assumes that the relationship between the dependent and each of the independent variables takes a particular form. This report was informed by logistic regression analysis - a method that summarises the relationship between a binary 'dependent' variable (one that takes the values '0' or '1') and one or more 'independent' explanatory variables. The tables in this annex show how the odds ratios for each category in significant explanatory variables compares to the odds ratio for the reference category (always taken to be 1.00).

3. Taking Model 1 (below) as an example, the dependent variable is whether people agree they regularly stop and speak to people in their area. If the respondent strongly agreed or agreed that they regularly stop and speak to people in their area, the dependent variable takes a value of 1. If they neither agreed nor disagreed, disagreed or disagreed strongly, it takes a value of 0. An odds ratio of above 1 means that, compared with respondents in the reference category, respondents in that category have higher odds of strongly agreeing or agreeing that they regularly stop and speak to people in their local area. Conversely, an odds ratio of below 1 means they have lower odds of saying this than respondents in the reference category. The 95% confidence intervals for these odds ratios are also important. Where the confidence interval does not include 1, this category is significantly different from the reference category. If we look at age in Model 1, we can see that those aged 65 or over an odds ratio of 3.62, indicating that they have higher odds of strongly agreeing or agreeing that they regularly stop and speak to people in their area compared with 18-29 year-olds (who are the reference category). The 95% confidence interval for this age group (2.08-6.28) does not include 1, indicating this difference is significant.

4. The significance of each independent variable is indicated by 'P'. A p-value of 0.05 or less indicates that there is less than a 5% chance we would have found these differences between the categories just by chance if in fact no such difference exists, while a p-value of 0.01 or less indicates that there is a less than 1% chance. P-values of 0.05 or less are generally considered to indicate that the difference is highly statistically significant, while a p-value of 0.06 to 0.10 may be considered marginally significant. The models below show the odds ratios and 95% confidence intervals for those variables with P-values of less than or equal to 0.10 only; other variables are simply listed with their P-values.

5. The models below were produced using the Complex Survey command (CS Logistic) in PASW. Unlike forward stepwise models, CS Logistic models can account for complex sample designs (in particular, the effects of clustering and associated weighting) when calculating odds ratios and determining significance.

Model 1: Factors associated with strongly agreeing or agreeing 'I regularly stop and speak to people in my area' (SSA 2009)

Dependent variable encoding
1 = Strongly agree/agree

0 = Neither/disagree/strongly disagree
Odds ratio 95% confidence interval
Age (p = 0.000)
18-29 (reference) 1.00
30-39 1.06 0.59-1.89
40-64 1.66 1.06-2.60
65+ 3.62 2.08-6.30
Urban-rural (p=0.000)
Large urban 1.00
Other urban 0.93 0.66-1.31
Accessible small town 1.52 0.77-3.01
Remote small town 2.63 1.50-4.61
Accessible rural 1.59 1.08-2.36
Remote rural 5.43 2.98-9.88
Highest educational qualification (p = 0.039)
No qualification (reference) 1.00
Degree or higher education 1.07 0.62-1.86
Highers or equivalent 1.82 0.94-3.52
Standard grade or other school level qualification 1.43 0.85-2.43
Whether school age children in household (p=0.057)
Yes (reference) 1.00
No 0.64 0.41-1.01
Gender (p=0.093)
Male (reference) 1.00
Female 1.31 0.95-1.79
Tenure (p=0.138)
NS-SEC socio-economic group (p = 0.148)
Whether has disability (p=0.518)
Income (p=0.542)
Area deprivation (p = 0.544)
Marital status (p = 0.762)

Nagelkerke R2 = 14.1%

Model 2: Factors associated with strongly agreeing or agreeing 'If my home was empty, I could count on one of my neighbours to keep an eye on it' (SSA 2009)

Dependent variable encoding
1 = Strongly agree/agree

0 = Neither/disagree/strongly disagree
Odds ratio 95% confidence interval
Tenure (p = 0.000)
Social renter (reference) 1.00
Owner 2.90 1.67-5.03
Private renter 0.56 0.29-1.07
Age (p=0.104)
Urban-rural (p=0.155)
Whether school age children in household (p=0.165)
Sex (p = 0.212)
Highest educational qualification (p=0.244)
Whether has disability (p=0.450)
Area deprivation (p = 0.627)
Marital status (p = 0.735)
NS-SEC socio-economic group (p = 0.868)
Income (p=0.975)

Nagelkerke R2 = 15.8%

Model 3: Factors associated with strongly agreeing or agreeing 'I feel that there are people in this area I could turn to for advice and support' (SSA 2009)

Dependent variable encoding
1 = Strongly agree/agree

0 = Neither/disagree/strongly disagree
Odds ratio 95% confidence interval
Urban-rural (p=0.001)
Large urban 1.00
Other urban 0.97 0.67-1.36
Accessible small town 1.21 0.74-1.98
Remote small town 2.04 1.21-3.43
Accessible rural 1.70 1.06-2.70
Remote rural 4.24 2.10-8.56
Area deprivation (p=0.007)
5th - Most deprived 1.00
4th 1.07 0.64-1.80
3rd 1.19 0.70-2.06
2nd 1.36 0.78-2.38
1st - Least deprived 2.25 1.28-3.96
Age (p = 0.029)
18-29 (reference) 1.00
30-39 0.77 0.45-1.32
40-64 0.93 0.55-1.58
65+ 1.76 0.85-3.63
Gender (p=0.085)
Male (reference) 1.00
Female 1.28 0.97-1.70
Tenure (p=0.089)
Social renter (reference) 1.00
Owner 1.18 0.65-2.13
Private renter 0.67 0.36-1.23
Whether school age children in household (p=0.182)
Whether has disability (p=0.216)
NS-SEC socio-economic group (p = 0.266)
Highest educational qualification (p=0.383)
Income (p=0.515)
Marital status (p = 0.981)

Nagelkerke R2 = 10.2%

Model 4: Factors associated with saying that 'most people can be trusted' rather than saying 'you can't be too careful in dealing with people' (SSA 2009)

Dependent variable encoding
1 = Most people can be trusted

0 = You can't bee too careful dealing with people/Don't know/Not answered
Odds ratio 95% confidence interval
Gender (p = 0.000)
Male (reference) 1.00
Female 0.61 0.47-0.79
Area deprivation (p=0.003)
5th - Most deprived 1.00
4th 1.20 0.75-1.94
3rd 1.61 1.00-2.61
2nd 2.00 1.21-3.32
1st - Least deprived 2.60 1.55-4.36
Urban-rural (p=0.006)
Large urban 1.00
Other urban 0.84 0.58-1.30
Accessible small town 1.36 0.62-3.00
Remote small town 1.78 1.18-2.68
Accessible rural 1.43 0.89-2.31
Remote rural 2.23 1.23-4.02
Highest educational qualification (p = 0.012)
No qualification (reference) 1.00
Degree or higher education 1.34 0.90-1.99
Highers or equivalent 1.30 0.82-2.07
Standard grade or other school level qualification 0.72 0.47-1.10
Tenure (p=0.079)
Social renter (reference) 1.00
Owner 1.56 1.06-2.30
Private renter 1.27 0.75-2.16
NS-SEC socio-economic group (p = 0.277)
Income (p=0.450)
Age (p = 0.475)
Whether school age children in household (p=0.669)
Whether has disability (p=0.963)
Marital status (p = 0.963)

Nagelkerke R2 = 16.3%

Model 5: Factors associated with having volunteered in the last 12 months (SHS 2010)

Dependent variable encoding
1 = Volunteered in last 12 months

0 = NOT volunteered in last 12 months
Odds ratio 95% confidence interval
Highest educational qualification (p = 0.000)
No qualification (reference) 1.00
Degree or higher education 2.98 2.17-4.08
Highers or equivalent 2.57 1.82-3.63
Standard grade or other school level qualification 1.82 1.30-2.55
Class (p=0.000)
Employers/mgrs & professional (reference) 1.00
Intermediate occupations 0.65 0.49-0.86
Small employers/ own account workers 0.87 0.63-1.21
Lower supervisory & technical 0.54 0.38-0.75
Semi-routine & routine occupations 0.61 0.47-0.79
Urban-rural (p=0.000)
Large urban (reference) 1.00
Other urban 1.17 0.93-1.46
Accessible small town 1.18 0.86-1.61
Remote small town 1.43 0.95-2.14
Accessible rural 1.61 1.21-2.14
Remote rural 2.68 1.86-3.87
Gender (p = 0.003)
Male (reference) 1.00
Female 1.32 1.10-1.58
Area deprivation (p=0.004)
5th - Most deprived (reference) 1.00
4th 1.57 1.13-2.18
3rd 1.38 0.99-1.93
2nd 1.71 1.22-2.39
1st - Least deprived 1.87 1.34-2.62
Income (p=0.067)
£11,999 or less (reference) 1.00
£12k-£22,999 0.74 0.55-1.01
£23K-£37,999 0.94 0.68-1.30
£38K+ 1.00 0.70-1.43
Age (p = 0.101)
Tenure (p=0.151)
Marital status (p = 0.375)
Whether has disability (p=0.643)

Nagelkerke R2 = 12.5%

Model 6: Factors associated with having done something active to register views rather than having done something passive or nothing at all to register views' (SSA 2009)

Dependent variable encoding
1 = Done something active to register views

0 = NOT done something active to register views
Odds ratio 95% confidence interval
Highest educational qualification (p = 0.000)
No qualification (reference) 1.00
Degree or higher education 2.98 1.95-4.56
Highers or equivalent 1.56 1.02-2.39
Standard grade or other school level qualification 1.41 0.95-2.08
Age (p = 0.007)
18-29 (reference) 1.00
30-39 0.71 0.42-1.20
40-64 1.32 0.80-2.19
65+ 1.57 0.81-3.04
Whether has disability (p=0.009)
Yes (reference) 1.00
No 0.70 0.54-0.91
Income (p=0.100)
£11,999 or less (reference) 1.00
£12k-£22,999 1.03 0.71-1.49
£23K-£37,999 0.75 0.47-1.20
£38K+ 0.88 0.51-1.50
Whether school age children in household (p=0.166)
Marital status (p = 0.277)
Area deprivation (p=0.325)
Urban-rural (p=0.449)
NS-SEC socio-economic group (p = 0.602)
Tenure (p=0.693)

Nagelkerke R2 = 11.2%

Model 7: Factors associated with strongly disagreeing or disagreeing that 'It is just too difficult for someone like me to do much about improving my local area' (SSA 2009)

Dependent variable encoding
1 = Strongly disagree/disagree

0 = Neither/agree/strongly agree
Odds ratio 95% confidence interval
Age (p = 0.000)
18-29 (reference) 1.00
30-39 2.47 1.44-4.22
40-64 4.05 2.48-6.63
65+ 3.13 1.70-5.76
Highest educational qualification (p = 0.000)
No qualification (reference) 1.00
Degree or higher education 2.50 1.39-4.50
Highers or equivalent 1.72 0.68-3.45
Standard grade or other school level qualification 1.00 0.53-1.90
Class (p=0.004)
Employers/mgrs & professional (reference) 1.00
Intermediate occupations 0.45 0.26-0.76
Small employers/ own account workers 0.58 0.33-1.01
Lower supervisory & technical 0.45 0.22-0.94
Semi-routine & routine occupations 0.66 0.43-1.00
Urban-rural (p=0.009)
Large urban 1.00
Other urban 0.98 0.67-1.44
Accessible small town 1.28 0.75-2.19
Remote small town 0.80 0.34-1.89
Accessible rural 1.04 0.62-1.75
Remote rural 3.05 1.65-5.66
Area deprivation (p=0.011)
5th - Most deprived 1.00
4th 1.12 0.34-1.98
3rd 0.72 0.40-1.28
2nd 1.38 0.72-2.66
1st - Least deprived 1.26 0.67-2.35
Marital status (p = 0.122)
Income (p=0.146)
Gender (p = 0.252)
Whether school age children in household (p=0.735)
Whether has disability (p=0.775)
Tenure (p=0.790)

Nagelkerke R2 = 21.4%

Model 8: Factors associated with strongly agree or tend to agree that 'I can influence decisions affecting my local area' (SHS 2010)

Dependent variable encoding
1 = Strongly agree/tend to agree

0 = Neither/tend to disagree/strongly disagree
Odds ratio 95% confidence interval
Highest educational qualification (p = 0.000)
No qualification (reference) 1.00
Degree or higher education 1.48 1.15-1.91
Highers or equivalent 1.29 0.97-1.70
Standard grade or other school level qualification 0.85 0.64-1.12
Urban-rural (p=0.001)
Large urban 1.00
Other urban 0.78 0.64-0.94
Accessible small town 0.95 0.72-1.25
Remote small town 0.67 0.46-0.96
Accessible rural 0.96 0.75-1.23
Remote rural 1.38 1.02-1.87
Age (p=0.203)
Whether has disability (p=0.278)
Tenure (p=0.416)
Gender (p = 0.425)
Income (p=0.591)
Area deprivation (p=0.686)
Class (p=0.737)
Marital status (p = 0.947)

Nagelkerke R2 = 2.9%

Model 9: Factors associated with strongly agreeing or tending to agree that 'I would like to be more involved in the decisions my council makes that affect my local area' (SHS 2010)

Dependent variable encoding
1 = Strongly agree/tend to agree
0 = Neither/tend to disagree/strongly disagree
Odds ratio 95% confidence interval
Age (p = 0.000)
18-29 (reference) 1.00
30-39 1.40 1.22-1.75
40-64 1.18 0.95-1.46
65+ 0.70 0.49-1.01
Highest educational qualification (p = 0.000)
No qualification (reference) 1.00
Degree or higher education 2.26 1.81-2.82
Highers or equivalent 1.71 1.34-2.18
Standard grade or other school level qualification 1.59 1.26-1.99
Urban-rural (p=0.007
Large urban 1.00
Other urban 0.78 0.66-0.92
Accessible small town 0.66 0.52-0.84
Remote small town 0.81 0.60-1.09
Accessible rural 0.78 0.63-0.97
Remote rural 0.84 0.63-1.11
Area deprivation (p=0.080)
5th - Most deprived 1.00
4th 1.03 0.82-1.29
3rd 1.10 0.87-1.39
2nd 0.87 0.69-1.11
1st - Least deprived 0.85 0.67-1.09
Whether has disability (p=0.314)
Tenure (p=0.353)
Marital status (p = 0.373)
Class (p=0.624)
Gender (p = 0.643)
Income (p=0.938)

Nagelkerke R2 = 4.9%

Model 10: Factors associated with anti-social behaviour being not a very big, or not a problem at all in local area (SSA 2009)

Dependent variable encoding
1 = Not a very big problem/not a problem at all
0 = Very big/quite a big problem
Odds ratio 95% confidence interval
Area deprivation (p=0.000)
5th - Most deprived 1.00
4th 1.70 1.06-2.74
3rd 2.21 1.26-3.86
2nd 4.58 2.71-7.75
1st - Least deprived 6.54 3.38-12.64
Urban-rural (p=0.001)
Large urban 1.00
Other urban 1.00 0.63-1.59
Accessible small town 0.62 0.41-0.95
Remote small town 0.99 0.53-1.85
Accessible rural 1.46 0.80-2.66
Remote rural 2.23 1.20-4.13
Income (p=0.014)
£11,999 or less (reference) 1.00
£12,000-22,999 1.34 0.84-2.14
£23,000-£37,999 1.34 0.82-2.20
£38,000+ 1.83 1.13-2.97
Class (p=0.061)
Employers/mgrs & professional (reference) 1.00
Intermediate occupations 0.86 0.52-1.41
Small employers/ own account workers 0.65 0.36-1.20
Lower supervisory & technical 1.45 0.81-2.59
Semi-routine & routine occupations 0.84 0.56-1.27
Age (p = 0.072)
18-29 (reference) 1.00
30-39 1.48 0.90-2.42
40-64 2.12 1.16-3.87
65+ 2.65 1.25-5.63
Marital status (p = 0.283)
Whether has disability (p=0.325)
Whether school age children in household (p=0.469)
Gender (p = 0.501)
Tenure (p=0.720)
Highest educational qualification (p=0.817)

Nagelkerke R2 = 20.1%


Contact

Email: Linzie Liddell