Hypothesis: The effect of ever having a paid job on satisfaction with the economy is stronger among residents who live in urban areas than among those who live in rural areas.

fre pdjobev.
RECODE pdjobev (1=1) (2=0) INTO paidjob_dummy.
VARIABLE LABELS paidjob_dummy ‘ever had a paid job’.
VALUE LABELS paidjob_dummy 0’no’ 1’yes’.
FREQUENCIES pdjobev paidjob_dummy.

fre stfeco.
RECODE stfeco (0 thru 5=0) (6 thru 10=1) INTO stfeco_dummy.
VARIABLE LABELS stfeco_dummy ‘satisfaction with economy’.
VALUE LABELS stfeco_dummy 1 ‘satisfied’ 0 ‘dissatisfied’.
FREQUENCIES stfeco stfeco_dummy.

FREQUENCIES domicil.
RECODE domicil (1 thru 2=1) (3 thru 5=0) INTO domicil_dummy.
VARIABLE LABELS domicil_dummy ‘big city or no’.
VALUE LABELS domicil_dummy 1’big city’ 0’not a big city’.
FREQUENCIES domicil domicil_dummy.

CROSSTABS stfeco_dummy BY paidjob_dummy BY domicil_dummy /CELLS=COLUMN /STATISTICS=CHISQ LAMBDA RISK.

Epsilon (not big city):19.4-22.2=-2.8
Epsilon (big city): 17.2-20.9=-3.7

In rural areas, the proportion of those who are satisfied with the economy is 2,8 percentage points lower among those who never had a paid job (than among those who have ever had a paid job.)
In urban areas, the proportion of those who are satisfied with the economy is 3,7 percentage points lower among those who have never had a paid job than among those who have ever had a paid job.

The difference of epsilons is 1,1, (because 1,1 is the difference between -2,8 and -3,7) –> the difference between those having a paid job and those not having a paid job is 1,1 percentage points lower in small cities than in big cities. Note: The difference is calculated with absolute values, so we basically subtract 3,7-2,8. The + or – sign only shows the direction of the effect. The size of the effect is not influenced by the sign of the effect. Here we only compare the effect sizes (3,7-2,8=1,1), not the direction.

In urban areas, the odds of being dissatisfied with the economy among people who never had a paid job is 1.268 times as high as among those who have ever had a paid job.
In rural areas, the odds of being dissatisfied with the economy among those who have never had a paid job is 1.182 times as high as those who have ever had a paid job.
The second-order odds ratio is 0.93 (1,183/1,268=0,93). The effect of ever having a paid job on satisfaction with the economic situation in big cities is 0.93 times as low as in other areas (not big cities).

We interpret the second-order odds ratio by comparing the relationship
between ever having a paid job and satisfaction with the economy among big city residents and the same among not big city residents.

Conclusion: The Chi-square test shows that the results are statistically not significant. Note: If the results were significant you should also examine the epsilons and/or the odds ratios to conclude your findings. Thus, there is no evidence for the interaction effect. It seems that the effect of domicile does not moderate the relationship between satisfaction with the economy and ever having a paid job. We say that it does not moderate/there is no interaction effect because we see that both the p values in the Chi-square table are not significant.

The results do not support the hypothesis stating that: “The effect of ever having a paid job on satisfaction with the economy is higher among residents who live in urban areas than among those who live in rural areas.” It seems that there is no interaction effect since both the p values are higher than 0,05.