Table 3

Initial exploratory regression showing only the variables which passed the cut-off p value

Independent variableOrdinary least squares estimation number of observations: 286
R-squared: 0.532
Akaike info: 2105.81
Jarque-Bera prob: 0.000
Summary of variable significance among all possible variable combinations
Coefficientt-statisticP value% Significant% Negative% Positive
WPS recipients—Veterans20.3742.3120.02273.05%0%100%
No central heating−28.46−2.9310.00470.72%80.88%19.12%
SIMD14.8091.6050.1118.41%6.13%93.87%
AFPS recipients—Veterans11.5021.1170.26550%0.67%99.33%
AFCS recipients—Veterans6.7491.1110.26725%6.25%93.75%
No qualifications−10.835−0.9980.3197.70%78.56%21.44%
Unemployment28.1870.8440.39949.41%16.24%83.76%
Alcohol−6.504−0.8020.4247.35%81.20%18.80%
  • R-squared: indicates how much variation of a dependent variable is explained by the independent variable(s).

  • Akaike info: estimates the relative amount of information lost by a given model: the less information a model loses, the higher the quality of that model.

  • Jarque-Bera prob: indicates if the data have a normal distribution. If it is far from zero, it signals the data do not have a normal distribution.

  • Coefficient: indicates the change in the dependent variable for one unit of change in the independent variable. A negative coefficient suggests that as the independent variable increases, the dependent variable tends to decrease.

  • t-statistic: the coefficient divided by its standard error, thus estimating the likelihood that the regression coefficient is different from zero.

  • p value: estimates what the odds are of the results to have happened. The lower it is, the less likely the results could have happened due to random chance.

  • AFCS, Armed Forces Compensation Scheme; SIMD, Scottish Index of Multiple Deprivation; WPS, War Pension Scheme.