The Effect of an Economic Crisis on Peoples's Health - Statistics Project Example

Q.No. 18 Means of health care expenditure (hse) vs. unemployed Means for variable health care expenditure unemployed Mean n Std. Deviation 0 437.971822350 784.83390 1 642.5993 710 756.93304 Total 444.2721 23060 784.77033 Interpretation: It is interpreted that the hse mean for unemployed with code 0 is 437.97 and the hse mean with unemployed code 1 is 642.6. The hse mean for unemployed with code 1 is greater than the hse mean for unemployed with code 0. The mean difference is 204.6 which seems to be a great difference . Q.No.19 Correlations Variable hh_income unemployed Pearson Correlation –0.177** Sig. (2–tailed) 0.000 n 18480 **. Correlation is significant at the 0.01 level (2–tailed). Interpretation: Out of 23060 cases, only 18480 cases have both unemployed and hh_income as non–empty cases. The correlation coefficient between unemployed and hh_income is negative ie. –0.177** which is highly significant because the probability of significance is 0.000 < 0.05, but as in magnitude, the correlation is weak negatives, but the no. of cases is 18,480, hence even this less correlation coefficient becomes significant. The significance of correlation is measured by t=r*sqrt(n–2)/1–r2, and the t value is 24.45 with probability 0.000. Hence the correlation despite being weak negative, becomes significant. This indicates that the unemployed status is negatively correlated with hh_income. Q.No.20 Between–Subjects Factors n hh_income 1 1554 2 1434 3 1751 4 1820 5 1817 6 1923 7 2020 8 2008 9 2061 10 2092 Tests of Between–Subjects Effects Dependent Variable: health care expenditure Source Type III Sum of Squares df Mean Square F Sig. Corrected Model 4.242E7a 9 4712796.908 13.259 0.000 Intercept 3.680E9 1 3.680E9 1.035E4 0.000 hh_income 4.242E7 9 4712796.908 13.259 0.000 Error 6.565E9 18470 355445.763 Total 1.025E10 18480 Corrected Total 6.607E9 18479 a. R Squared = 0.006 (Adjusted R Squared = .006) Parameter Estimates Dependent Variable:health care expenditure Parameter B Std. Error t Sig. 95% Confidence Interval Lower Bound Upper Bound Intercept 395.509 13.035 30.342 0.000 369.959 421.058 [hh_income=1] 149.913 19.966 7.508 0.000 110.778 189.048 [hh_income=2] 127.813 20.440 6.253 0.000 87.750 167.877 [hh_income=3] 80.534 19.311 4.170 0.000 42.683 118.384 [hh_income=4] 75.294 19.110 3.940 0.000 37.836 112.752 [hh_income=5] 25.883 19.119 1.354 0.176 –11.592 63.357 [hh_income=6] 9.593 18.835 .509 0.611 –27.325 46.510 [hh_income=7] 24.022 18.598 1.292 0.196 –12.431 60.475 [hh_income=8] 42.993 18.626 2.308 0.021 6.485 79.502 [hh_income=9] 2.944 18.503 .159 0.874 –33.324 39.212 [hh_income=10] 0 . . . . . a. This parameter is set to zero because it is redundant. Interpretation: Since house hold income is having limited values from 1 to 9, it is better to treat it as dummy variable rather than a real variable). Actually hh income is a categorical variable because for the 23060 cases it is limited from 1 to 9. So it can be treated as categorical variable rather than continuous variable. The lowest income category is hh–income=1. The predicated level of health care expenditure is 395.509+149.913*1=545.422. The highest income category is 9 for which the predicted health care expenditure is 2.944*9+395.509=422.005. The confidence interval of some categories although include negative values for the lower values, yet we can treat them as 0. The confidence interval throws a useful information as the house hold income level the health care expenditure decreases. This naturally happens because when a house hold income is more, the persons feel more motivated and feels happy and naturally emotional happiness increases. This may be the cause for decrease in health care expenditure. Q.21 Estimated Distribution Parameters hce Normal Distribution Location 444.27 Scale 784.77 Interpretation: From the above QQ plots and Graphs Legacy dialogs it is observed that the straight line in the plot represents expected values when the data are normally distributed. The observed values deviate markedly from that line, especially as hce increases. The Legacy dialogs also shows that some of the outliers clearly visible, but the outliers are within a certain limits. The graph also shows that some homoscedasticity present in the data. Q22. The histogram of the variable ln_hce is given below: The mean value of ln_hce is 5.89 with SD 1.002 with 17,432 valid cases. One–Sample Kolmogorov–Smirnov Test – ln_hce n 17432 Normal Parameters Mean 5.8880 Std. Deviation 1.00196 Most Extreme Differences Absolute 0.107 Positive 0.107 Negative –0.074 Kolmogorov–Smirnov Z 14.184 Asymp. Sig. (2–tailed) 0.000 Interpretation: From the above table it is inferred that the distribution of ln_hce is not normal because the Komogorov–Smirov Z is 14.184>1.96 and the probability of significance is less than 0.05. Model Summary R R Square Adjusted R Square Std. Error of the Estimate 0.072* 0.005 0.005 0.99544 ANOVA Source of variation Sum of Squares df Mean Square F Sig. Regression 73.173 1 73.173 73.845 0.000** Residual 13931.041 14059 0.991 Total 14004.214 14060 a. Predictors: (Constant), hh_income, b. Dependent Variable: ln_hce Coefficients Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta (Constant) 6.043 0.019 316.760 0.000 hh_income –0.025 0.003 –0.072 –8.593 0.000 Dependent Variable: ln_hce From the above table of regression, the regression coefficient is highly significant since the F value is 73.845 with probability of significance 0.000 ( Read More