Quantitative Analysis of Social Research Data using SPSS - Essay Example

Quantitative Analysis of Social Research Data using SPSS Question Spain, United Kingdom and Portugal were considered forthis analysis. Since these countries have entirely different existences and locations, it is interesting to know how men and women perceive happiness in these countries. Question 2 A categorical variable named agecat was generated from continuous age and the categories labelled using the syntax provided in appendix 1. Question 3 A description of the mean happiness quotient by gender was conducted for each country and the results are presented below. T tests were used to statistically assess the difference in means between men and women in each country. Since we are comparing means between two groups, it’s appropriate to use a t test. We assume that the variances of happiness quotients are unknown and so they are estimated from the data. Difference in means for men and women in Spain Looking at the average happy quotient by gender in table 1, the mean happy quotient for males is 7.719201 whereas for females is 7.550874. Table 1: Average happy quotient by gender for Spain Gender N Mean Std. Deviation Male 901 7.719 1.569 Female 973 7.551 1.738 A t test to compare means in the two groups was conducted and the results in table 2 show that the hypothesis of equal means is rejected and therefore the mean for men and women are not equal (Pagano, 2006). While conducting this test, we assume that the variances are not equal. Table 2: T test for equality of means for Spain t df Sig. (2-tailed) Std. Error Difference         2.203 1870.797 0.027 0.076 Difference in means for men and women in United Kingdom The average happy quotient by gender for United Kingdom is presented in table 3. From the results, it can be seen that the means for men and women are approximately 7.4 Table 3: Average happy quotient by gender for United Kingdom Gender N Mean Std. Deviation Male 1078 7.44 1.94 Female 1315 7.42 1.96 A t test to compare means in the two groups was conducted and the results in table 4 show that the hypothesis of equal means is not rejected at 5% level of significance. Therefore we cannot conclude that the mean for men and women are not equal (Pagano, 2006). While conducting this test, we assume that the variances are equal. Table 4: T test for equality of means for United Kingdom t df Sig. (2-tailed) Std. Error Difference 0.226 2391 0.82 0.08 Difference in means for men and women in Portugal Assumption considered is that variances are not equal. Table 5 shows the average happy quotient by gender for Portugal. The means for men and women are 6.643 and 6.286 respectively. A t test to compare means in the two groups was conducted and the results in table 6 show that the mean for men and women are not equal since the hypothesis of equal means is rejected at 5% level of significance. Table 5: Average happy quotient by gender for Portugal Gender N Mean Std. Deviation Male 848 6.643 1.739 Female 1326 6.286 1.934 Table 6: T test for equality of means for Portugal t df Sig. (2-tailed) Std. Error Difference 4.479 1940.528 0.00 0.079 Question 4 A description of the mean happiness quotient by age group was conducted for Spain, United Kingdom and Portugal. The results are presented below. Analysis of variance was done to statistically assess the difference in means across the age groups in each country. ANOVA assumes that the age groups are independent of each other. Difference in means between age groups in Spain Table 7 presents the mean happiness quotient for each age group in Spain. The respondents aged 35 years and below had the highest mean happiness quotient and respondents aged 55 years and above had the least happiness quotient. Table 8: Average happiness quotient by age group for Spain Age group N Mean Std. Deviation Std. Error under 35 years 651 7.919 1.424 0.056 35 to 54 years 637 7.584 1.691 0.067 55 and above 586 7.365 1.817 0.075 Total 1874 7.632 1.661 0.038 Analysis of variance was done to statistically assess the difference in means across the age groups in Spain. Results presented in table 8 show that the P value is less than 0.05 so we reject the null hypothesis. Therefore the mean happiness quotient among different age groups is different. (Field, 2009) Table 8: Analysis of variance for Spain Sum of Squares df Mean Square F Sig. Between Groups 96.652 2 48.326 17.836 0.00 Within Groups 5069.291 1871 2.709 Total 5165.944 1873 Difference in means between age groups in United Kingdom Table 9 shows the mean happiness quotient for each age group in United Kingdom. The respondents aged 55years and above had the highest mean happiness quotient and respondents aged between 35 and 54 years had the least happiness quotient. Table 9: Average happiness quotient by age group for United Kingdom Age group N Mean Std. Deviation Std. Error under 35 years 656 7.416 1.822 0.071 35 to 54 years 765 7.244 2.007 0.073 55 and above 972 7.592 1.980 0.063 Total 2393 7.433 1.951 0.040 ANOVA was conducted to statistically assess the difference in means across the age groups in United Kingdom. Results are presented in table 10 and they show that the hypothesis of equal means is rejected at 5% level of significance. Therefore the mean happiness quotient among different age groups is different (Ciaran, et al., 2009). Table 10: Analysis of variance for United Kingdom Sum of Squares df Mean Square F Sig. Between Groups 51.822 2 25.911 6.837 0.001 Within Groups 9057.528 2390 3.790 Total 9109.351 2392 Difference in means between age groups in Portugal Table 11 presents the mean happiness quotient for each age group in Portugal. The respondents aged 35years and below had the highest mean happiness quotient and respondents aged 55 years and above had the least happiness quotient. Table 11: Average happiness quotient by age group for Portugal Age group N Mean Std. Deviation Std. Error under 35 years 585 7.003 1.639 0.068 35 to 54 years 627 6.671 1.819 0.073 55 and above 962 5.914 1.898 0.061 Total 2174 6.425 1.868 0.040 Analysis of variance was done to statistically assess the difference in means across the age groups in Portugal. Results presented in table 12 show that the hypothesis of equal means is rejected at 5% level of significance. Therefore it can be concluded that the mean happiness quotient among different age groups is different. Table 12: Analysis of variance for Portugal Sum of Squares df Mean Square F Sig. Between Groups 485.277 2 242.639 74.191 0.000 Within Groups 7100.151 2171 3.270 Total 7585.428 2173 Question 5 Three categories were created for a new variable called labour market position. The categories are; “those in paid employment”, “the retired” and “not active in the labour market”. The category paid employment consists of those in paid work, community or military service whereas the retired category consists of those who are retired. The rest of the categories were recoded to not active in the labour market. The syntax used to create the variable is presented in appendix 4 (Field, 2009). Question 6 Cross tabulations were used to assess association between gender and labour market position. A chi square test was used to assess the significance of the association and the results presented below. The chi-square test assumes that there are two or more groups in each variable and that the variables are categorical. Association between gender and labour market position for Spain A cross tabulation between gender and labour market position in Spain is presented in Table 13 below. There are many male respondents belonging to the paid employment category whereas most female respondents belong to not active in the labour market category. Table 13: Cross tabulation of gender and labour market position for Spain male female Total those in paid employment 584 436 1020 the retired 177 58 235 not active in the labour market 141 480 621 Total 902 974 1876 A chi-square test was performed to assess the significance of the association and the results are presented in table 14. The association between gender and labour market position in Spain is highly significant at 5% level of significance (Field, 2009). Table 14: Chi square test for Spain Value df Asymp. Sig. (2-sided) Pearson Chi-Square 264.4182 2 0.00 Association between gender and labour market position for United Kingdom Table 15 below shows a cross tabulation between gender and labour market position in United Kingdom. Most male as well as female respondents belong to those in paid employment category. Table 15: Cross tabulation of gender and labour market position for United Kingdom male female Total those in paid employment 641 592 1233 the retired 238 382 620 not active in the labour market 200 341 541 Total 1079 1315 2394 A chi-square test was performed to assess the significance of the association and the results are presented in table 16. The association between gender and labour market position in United Kingdom is highly significant at 5% level of significance. Table 16: Chi-square test for United Kingdom Value df Asymp. Sig. (2-sided) Pearson Chi-Square 49.35587 2 0.00 Association between gender and labour market position for Portugal Table 17 below shows a cross tabulation between gender and labour market position in Portugal. Most male as well as female respondents belong to those in paid employment category. For males, the least number of respondents were in the not active in the labour market category whereas for females the least number of respondents were in the retired category. Table 17: Cross tabulation of gender and labour market position for Portugal male female Total those in paid employment 458 570 1028 the retired 271 383 654 not active in the labour market 134 406 540 Total 863 1359 2222 A chi-square test was performed to assess the significance of the association and the results are presented in table 18. The association between gender and labour market position in Portugal is highly significant at 5% level of significance. Table 18: Chi-square test for Portugal Value df Asymp. Sig. (2-sided) Pearson Chi-Square 60.69629 2 0.000 Question 7 Correlation matrix for house size, age, religiosity and Happy for males shows that variables that are highly associated are: age and house size, age and religiosity, happy with house size and age (Lewis-Beck et.al., 1995). Table 19: Correlation matrix for household size, age, religiosity and Happiness for males House size Age Religiosity Happy House size Correlation Coefficient 1 -0.399 -0.019 0.122   Sig. (2-tailed) 0.000 0.312 0.000   N 2838 2836 2789 2821 Age Correlation Coefficient -0.399 1 0.239 -0.071   Sig. (2-tailed) 0.000 0.000 0.000   N 2836 2842 2793 2825 Religiosity Correlation Coefficient -0.019 0.239 1 0.015   Sig. (2-tailed) 0.312 0.000 0.433   N 2789 2793 2795 2784 Happy Correlation Coefficient 0.122 -0.071 0.015 1   Sig. (2-tailed) 0.000 0.000 0.433   N 2821 2825 2784 2827 Correlation matrix for house size, age, religiosity and Happy for females shows that variables that are highly associated are: age and house size, religiosity with house size, religiosity and age, happy with house size and age. Table 20: Correlation matrix for household size, age, religiosity and Happiness for females House size Age Religiosity Happy House size Correlation Coefficient 1 -0.521 -0.113 0.154   Sig. (2-tailed) - 0.000 0.000 0.000   N 3646 3641 3603 3612 Age Correlation Coefficient -0.521 1 0.313 -0.177   Sig. (2-tailed) 0.000 0.000 0.000   N 3641 3643 3599 3609 Religiosity Correlation Coefficient -0.113 0.313 1 -0.057   Sig. (2-tailed) 0.000 0.000 0.001   N 3603 3599 3604 3585 Happy Correlation Coefficient 0.154 -0.177 -0.057 1   Sig. (2-tailed) 0.000 0.000 0.001   N 3612 3609 3585 3614 Three potential control variables identified for this analysis are age, household size and gender. These variables were highly related to happiness for females and males. Question 8 A linear regression analysis was carried out to examine factors that are significantly associated with happiness among men and women in each country and the results are presented in table 21 (Pagano, 2006). The results show that all the factors considered are highly associated with happiness Table 21: Linear regression analysis results Beta Std. Error t Sig. Intercept 8.616 0.139 62.031 0.000 Gender -0.149 0.046 -3.215 0.001 Age of respondent, calculated -0.008 0.001 -6.132 0.000 House hold size 0.117 0.020 5.960 0.000 country -0.560 0.029 -19.155 0.000 Question 9 Variable happy was classified into two categories: happy and unhappy. A logistic regression was fit to the data to examine factors that are related to being happy among men and women in the three countries. In this analysis male was used as the reference for gender, 35 years and below for age whereas Spain was used as reference for country (Fielding & Gilbert , 2006). Results in table 22 show that the odds of being happy versus being unhappy are 1.444 times higher for respondents aged between 35 and 54 years as compared to those aged 35 years and below. The odds of being happy versus being unhappy are 1.303 higher for females as compared to males, whereas the odds of being happy versus being unhappy are higher for respondents from United Kingdom and Portugal as compared to those from Spain. Table 22: Results of logistic regression B S.E. Wald df Sig. Exp(B) agecat 11.928 2 0.003 agecat(1) 0.367 0.133 7.623 1 0.006 1.444 agecat(2) -0.088 0.115 0.578 1 0.447 0.916 gndr(1) 0.264 0.097 7.364 1 0.007 1.303 country 54.578 2 0.000 country(1) 0.996 0.141 50.231 1 0.000 2.708 country(2) 0.433 0.103 17.707 1 0.000 1.542 hhmmb 0.311 0.047 44.215 1 0.000 1.365 Constant 1.139 0.122 87.223 1 0.000 3.123 Question 10 The results show that all the factors considered are highly associated with happiness Overall the men showed a higher chance of being happy as compared to women in the three countries. The results further reveal that the level of happiness varies in the three countries whereby respondents from United Kingdom showed higher odds of being happier than the ones from Spain and Portugal. For further analysis, it would be interesting to consider more countries as well as other factors that affect happiness of men and women. References Ciaran, A., Miller , R., Fullerton, D. & Maltby, J., 2009. SPSS for Social Scientists. Second Edition ed. s.l.:Palgrave Macmillan. Field, A., 2009. Discovering Statistics Using SPSS. 3rd Edition ed. s.l.:Sage. Field, A., 2009. Discovering Statistics using SPSS for Windows. 3rd Edition ed. s.l.:Sage. Fielding, J. & Gilbert , N., 2006. Understanding Social Statistics. Second Edition, ,. London: Sage Publications. Pagano, R. R., 2006. Understanding Statistics in the Behavioural Sciences. West – 8th Edition ed. s.l.:s.n. Lewis-Beck M. S. (1995). Data Analysis: An Introduction. (Sage University paper series on Quantitative Applications in Social Sciences, 01-103). Thousand Oaks, CA: Sage. Syntax used to obtain results from SPSS Appendix 1 Creating age into categories and labelling categories compute agecat=1. if (age>=35) & (age=55) agecat=3. execute. variable labels agecat "age in categories". value labels agecat 1"under 35 years" 2"35 to 54 years" 3"55 and above". execute. Appendix 2 Distribution of happy by gender and t test CROSSTABS /TABLES=happy BY gndr /FORMAT= AVALUE TABLES /CELLS= COUNT ROW /COUNT ROUND CELL /BARCHART . T-TEST GROUPS = gndr(1 2) /MISSING = ANALYSIS /VARIABLES = happy /CRITERIA = CI(.95) Appendix 3 Distribution of happiness quotient by age group and ANOVA ONEWAY happy BY agecat /STATISTICS DESCRIPTIVES /MISSING ANALYSIS . Appendix 4 # Creating variable called labour_market_position COMPUTE labour_market_position. if (mnactic=1) or (mnactic=7) labour_market_position=1. if (mnactic=6) labour_market_position=2. if (mnactic!= 1) or (mnactic!= 6) or (mnactic!= 7) labour_market_position=3. execute. variable labels labour_market_position "Labour market position". value labels agecat 1"those in paid employment" 2"the retired" 3"not active in the labour market". execute. Appendix 5 CROSSTABS /TABLES=labour_market_position BY gndr /FORMAT= AVALUE TABLES /STATISTIC=CHISQ /CELLS= COUNT /COUNT ROUND CELL . Appendix 5 Correlation Matrix NONPAR CORR /VARIABLES=hhmmb age rlgdgr happy /PRINT=SPEARMAN TWOTAIL NOSIG /MISSING=PAIRWISE . Appendix 6 Linear regression REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA /CRITERIA=PIN(.05) POUT(.10) /NOORIGIN /DEPENDENT happy /METHOD=ENTER gndr age hhmmb country Appendix 7 #categorising happiness compute binhappy. if (happy=0) or (happy=1) or (happy=2) or (happy=3) or (happy=4) binhappy=1. if (happy=5) or (happy=6) or (happy=7) or (happy=8) or (happy=9) or (happy=10) binhappy=2. execute. value labels binhappy 1"unhappy" 2"happy". execute. Appendix 8 Logistic regression LOGISTIC REGRESSION VARIABLES binhappy /METHOD = ENTER agecat gndr country hhmmb /CONTRAST (agecat)=Indicator /CONTRAST (gndr)=Indicator /CONTRAST (country)=Indicator /CRITERIA = PIN(.05) POUT(.10) ITERATE(20) CUT(.5) . Read More