Social Capital and Health - Findings from the Scottish Health Survey and Scottish Social Attitudes Survey

This paper explores the Scottish evidence for a link between social capital and health outcomes to inform the ongoing development of an assets-based approach to addressing health problems and inequalities.


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 self-assessed health. If the respondent rated their own health as fair, bad or very bad, the dependent variable takes a value of 1. If they rated their health as good or very good, 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 saying their health is fair/bad/very bad. 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 35-44 have an odds ratio of 2.32, indicating that they have higher odds of saying standards in the health service have fallen compared with 16-24 year-olds (who are the reference category). The 95% confidence interval for this age group (1.07-5.02) 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 rating own health as fair/bad/very bad rather than good/very good (SHeS 2009)

Dependent variable encoding
1 = Fair/bad/very bad
0 = Good/very good
Odds ratio 95% confidence interval Sample size (weighted)
Age (p = 0.002)
16-24 (reference) 1.00 241
24-34 1.96 0.81-4.76 360
35-44 2.32 1.07-5.02 432
45-54 3.42 1.50-7.80 444
55-64 4.00 1.75-9.14 373
65-74 3.29 1.35-8.02 272
75+ 6.01 2.31-15.61 204
Area deprivation (SIMD quintiles) (p < 0.001)
Most deprived (reference) 1.00 412
2nd 0.86 0.58-1.26 493
3rd 0.87 0.56-1.35 430
4th 0.48 0.32-0.72 471
Least deprived 0.34 0.21-0.54 519
Economic status (p < 0.001)
Paid employment, self-employed or govt. training (ref.) 1.00 1414
In education 0.70 0.22-2.25 94
Looking for/intending to look for work 1.50 0.81-2.76 71
Permanently unable to work 29.12 14.34-59.15 104
Retired 2.95 1.82-4.77 505
Looking after home/family 1.84 1.14-2.98 138
Income quartiles (p = 0.084)
Lowest quartile (reference) 1.00 393
2nd quartile 0.82 0.57-1.16 422
3rd quartile 0.63 0.42-0.96 583
Highest quartile 0.50 0.30-0.84 622
How involved feel in local community (p = 0.050)
A great deal (reference) 1.00 109
A fair amount 1.51 0.78-2.92 509
Not very much 1.75 0.91-3.38 1125
Not at all 2.30 1.15-4.58 582
Feel can influence decisions over local area (p = 0.016)
Agree/strongly agree (reference) 1.00 498
Neither 0.69 0.44-1.06 655
Disagree 1.08 0.71-1.65 933
Strongly disagree 1.41 0.77-2.56 241
Number of people could turn to in a crisis (p = 0.030)
2 or fewer (reference) 1.00 270
3 or 4 people 0.79 0.53-1.19 412
5 or 6 people 0.54 0.36-0.81 699
7-10 people 0.70 0.45-1.10 515
11+ people 0.61 0.38-0.97 430
Personal contact with family, friends & neighbours (p = 0.033)
Most days (reference) 1.00 1557
Once or twice a week 1.30 0.97-1.74 627
Once or twice a month or less 1.87 1.10-3.18 142
Sex (p = 0.328)
NS-SEC socio-economic group (p = 0.590)
Highest educational qualification (p = 0.104)
Marital status (p = 0.993)
General social trust (p = 0.638)

Nagelkerke R2 = 38.4%

Model 2: Factors associated with rating own health as fair/bad/very bad rather than good/very good (SSA 2009)

Dependent variable encoding
1 = Fair/bad/very bad
0 = Good/very good
Odds ratio 95% confidence interval Sample size (weighted)
Age (p < 0.001)
16-24 (reference) 1.00 128
24-34 3.33 1.07-10.36 183
35-44 5.38 1.92-15.10 238
45-54 10.09 3.61-28.23 214
55-64 9.74 3.19-29.73 188
65-74 8.85 2.67-29.38 155
75+ 11.58 3.31-40.45 74
Economic status (p < 0.001)
Paid employment, education or govt. training (ref.) 1.00 734
Unemployed 1.60 0.74-3.48 68
Permanently sick or disabled 48.61 13.53-174.75 55
Retired 2.23 1.19-4.19 250
Looking after home/family 2.66 1.61-4.40 73
Income quartiles (p = 0.025)
£11,999 or less (reference) 1.00 195
£12,000-£22,999 0.90 0.54-1.51 236
£23,000-£37,999 0.49 0.29-0.84 250
£38,000+ 0.50 0.24-1.04 315
Income unknown 1.03 0.51-2.08 183
Sex (p = 0.073)
Male (Reference) 1.00 560
Female 0.73 0.52-1.03 620
NS-SEC socio-economic group (p = 0.075)
Semi-routine or routine occupations (reference) 1.00 363
Lower supervisory & technical 0.87 0.47-1.62 143
Small employers & own account workers 0.67 0.31-1.44 101
Intermediate occupations 0.44 0.24-0.81 140
Employers, managers & professionals 0.48 0.26-0.88 433
Highest educational qualification (p = 0.100)
No qualification (reference) 1.00 223
Higher education or degree 0.74 0.39-1.39 401
Highers or equivalent 1.71 0.81-3.61 208
Standard grade or other school level qualification 0.82 0.54-1.24 347
Agree/disagree 'I feel that there are people in this area I could turn to for advice and support' (p = 0.026)
Agree strongly (reference) 1.00 279
Agree 0.62 0.35-1.08 575
Neither 1.49 0.71-3.13 134
Disagree/disagree strongly 1.19 0.59-2.39 192
General social trust (p = 0.090)
Most people can be trusted (reference) 1.00 626
Can't be too careful dealing with people 1.48 0.94-2.35 553
Area deprivation (SIMD quintiles) (p = 0.410)
Marital status (p = 0.566)
Agree/disagree too difficult for someone like me to do much about improving local area (p = 0.853)
Agree/disagree 'I regularly stop and speak to people in my area' (p = 0.259)

Nagelkerke R2 = 34.8%

Model 3: Factors associated with below average WEMWBS score (SHeS 2009)

Dependent variable encoding
1 = 1 standard deviation or more below average WEMWBS score
0 = Average/above average score
Odds ratio 95% confidence interval Sample size (weighted)
Age (p = 0.043)
16-24 (reference) 1.00 221
24-34 0.66 0.35-1.26 338
35-44 1.04 0.56-1.93 404
45-54 1.10 0.54-2.25 413
55-64 0.61 0.27-1.36 348
65-74 0.44 0.17-1.16 245
75+ 0.73 0.27-2.00 164
Economic status (p < 0.001)
Paid employment, self-employed or govt. training (ref.) 1.00 1325
In education 0.88 0.32-2.42 83
Looking for/intending to look for work 2.33 1.24-4.38 66
Permanently unable to work 9.86 5.52-17.62 91
Retired 2.03 1.10-3.77 441
Looking after home/family 1.53 0.83-2.80 127
Area deprivation (SIMD quintiles) (p = 0.077)
1st (most deprived) 1.00 366
2nd 1.48 0.96-2.27 458
3rd 1.28 0.80-2.06 395
4th 0.96 0.61-1.54 436
5th - most deprived 0.77 0.44-1.38 477
Marital status (p = 0.084)
Married/civil partnership (reference) 1.00 1176
Living as married 1.15 0.67-1.97 221
Single 1.68 1.06-2.67 428
Separated 1.82 0.87-3.77 50
Divorced/dissolved civil partnership 1.68 0.94-3.00 113
Widowed/surviving civil partner 1.61 0.93-2.78 144
How involved feel in local community (p = < 0.001)
A great deal (reference) 1.00 100
A fair amount 0.70 0.34-1.44 467
Not very much 0.93 0.47-1.85 1055
Not at all 1.79 0.89-3.66 509
Number of people could turn to in a crisis (p < 0.001)
2 or fewer (reference) 1.00 237
3 or 4 people 1.15 0.74-1.78 369
5 or 6 people 0.51 0.32-0.80 642
7-10 people 0.62 0.37-1.04 481
11+ people 0.49 0.29-0.85 403
Whether feel can influence decisions over local area (p = 0.008)
Agree/strongly agree (reference) 1.00 472
Neither 1.38 0.87-2.20 597
Disagree 1.99 1.26-3.16 839
Strongly disagree 2.31 1.33-4.01 224
Sex (p = 0.11)
Income quartiles (p = 0.544)
NS-SEC socio-economic group (p = 0.295)
Highest educational qualification (p = 0.158)
General social trust (p = 0.104)
How often have personal contact with family, friends or neighbours (p = 0.168)

Nagelkerke R2 = 25.1%

Contact

Email: Linzie Liddell

Back to top