Negative coefficient of year_dummy4 (-0.0607) with a significance of p = 0.043
indicates that social trust was lower in 2018 than in 2004. This possibly indicates a
break in the previous positive trend. Soocial trust was significantly higher in earlier
years like 2008 and 2016 compared to 2004 and there was a decline in social trust in
2018 compared to 2004.
The overall model explains only 0.65% of the variation in social trust, suggesting
other factors are more important.
Compared to Glaeser et al. (2000), my analysis shows a similar pattern of variation in
social trust over time. With significant differences between years. The low R²
suggests that year dummies alone explain very little of the variation in social trust,
But that is not a big issue. Although it is highlighting the importance of additional
predictors used by Glaeser et al.
Variable,Coefficient (SE),P-value
Intercept,2.3474 (0.0222),0.0
Year Dummy 1,0.1090 (0.0308),0.0
Year Dummy 2,0.0747 (0.0290),0.01
Year Dummy 3,0.0783 (0.0319),0.014
Year Dummy 4,-0.0607 (0.0300),0.043
Standard errors are in parentheses.
*Significance level: p < 0.05.
!IMPORTANT NOTE!:
I did not know well how to make a table, moreover I had not enough time left. I will try
my best for the coming weeks to make it better.
Appendix
gen year_dummy = year - 1998
tabulate year, generate(year_dummy)
replace age = 1 if age < 25
replace age = 2 if age >= 25 & age <= 34
replace age = 3 if age >= 35 & age <= 44,
replace age = 4 if age >= 45 & age <= 54
replace age = 5 if age >= 55
label define age_cat_label 1 "<25" 2 "25-34" 3 "35-44" 4 "45-
54" 5 "55+"
label define age 1 "<25" 2 "25-34" 3 "35-44" 4 "45-54" 5 "55+"
gen college_education = 0
replace college_education = 1 if degree
// dit betekent je scoort 1 op degree dummy wanneer je hoger
dan 2 scoort op degree en geen missing hebt op degree hebt. (!
missing staat voor GEEN MISSING)
tab degree, miss
gen degree_dummmy = (degree > 2 & !missing(degree))
replace degree_dummmy = . if degree == .a
tab degree_dummy, miss
label list
indicates that social trust was lower in 2018 than in 2004. This possibly indicates a
break in the previous positive trend. Soocial trust was significantly higher in earlier
years like 2008 and 2016 compared to 2004 and there was a decline in social trust in
2018 compared to 2004.
The overall model explains only 0.65% of the variation in social trust, suggesting
other factors are more important.
Compared to Glaeser et al. (2000), my analysis shows a similar pattern of variation in
social trust over time. With significant differences between years. The low R²
suggests that year dummies alone explain very little of the variation in social trust,
But that is not a big issue. Although it is highlighting the importance of additional
predictors used by Glaeser et al.
Variable,Coefficient (SE),P-value
Intercept,2.3474 (0.0222),0.0
Year Dummy 1,0.1090 (0.0308),0.0
Year Dummy 2,0.0747 (0.0290),0.01
Year Dummy 3,0.0783 (0.0319),0.014
Year Dummy 4,-0.0607 (0.0300),0.043
Standard errors are in parentheses.
*Significance level: p < 0.05.
!IMPORTANT NOTE!:
I did not know well how to make a table, moreover I had not enough time left. I will try
my best for the coming weeks to make it better.
Appendix
gen year_dummy = year - 1998
tabulate year, generate(year_dummy)
replace age = 1 if age < 25
replace age = 2 if age >= 25 & age <= 34
replace age = 3 if age >= 35 & age <= 44,
replace age = 4 if age >= 45 & age <= 54
replace age = 5 if age >= 55
label define age_cat_label 1 "<25" 2 "25-34" 3 "35-44" 4 "45-
54" 5 "55+"
label define age 1 "<25" 2 "25-34" 3 "35-44" 4 "45-54" 5 "55+"
gen college_education = 0
replace college_education = 1 if degree
// dit betekent je scoort 1 op degree dummy wanneer je hoger
dan 2 scoort op degree en geen missing hebt op degree hebt. (!
missing staat voor GEEN MISSING)
tab degree, miss
gen degree_dummmy = (degree > 2 & !missing(degree))
replace degree_dummmy = . if degree == .a
tab degree_dummy, miss
label list
