Assignment 2 Gintarė Biterytė (UČO: 477937) Ekaterina Volevach (UČO: 460236) In the second assignment we are changing our hypothesis and focusing not on divorces but on changes between marital statuses (“never married”, “married”, “divorced” and “widowed”) and preferred gender roles in family. We suppose that attitude toward gender roles changes alongside with a position that people are taking regarding the institute of family and marriage. H1: Married people have more traditional gender attitude than people that got divorced; H2: Widowed people have more traditional gender attitude than married. OK, this happens, but please (1) update your first assignment (into the final paper) and (2) explain more precisely, what the “traditional gender attitudes” are. Method of research: In order to answer the main research question and confirm or refute hypotheses, we used the data from ‘‘European Values Study 2008: Integrated Dataset’’. The first one is “Current legal marital status”. The second one was created by using several variables from dataset (v159, v160, v161, v162, v163, v164, v165, v166). All these variables show the respondent's egalitarian or traditional attitude towards gender roles, however, some of them have opposite polarity, therefore they need turning over. 1 step: Our first step was recoding variables v160, v161, v162 by using the command “recode v160 v161 v162 (1=4) (2=3) (3=2) (4=1). You should write sociological paper, not describe your statistical procedures J I will not take any points this time but, please, write in common language in the final paper (like: “We recoded all variables in order to have them identically oriented”) 2 step: Considering our analysis purposes, it is necessary that one variable include both dimensions of gender attitudes (egalitarian and traditional). Therefore, the second step was the index of attitudes to gender roles formation (“grolesind”). grolesind = v159+v160+v161+v162+v163+v164+v165+v166 I am interested what the codes mean! What is v159, v160, v161…? That’s what I wrote you in the previous step – write in common language… We had these eight indicators of gender attitudes: willingness to do household chores… (or whatever) 3 step: To understand on what values of the index another aspect begins, we tabulated it: Table 1. Gender roles attitude index. The table format tis wrong! Why do you show me all the frequencies, percents and cumulative? Choose one, maybe show better the histogram. This is unreadable. Grolesind Freq. Percent Cum. 8 401 0.75 0.75 9 501 0.93 1.68 10 760 1.41 3.09 11 1,443 2.68 5.78 12 1,838 3.42 9.20 13 2,733 5.09 14.28 14 3,532 6.57 20.85 15 4,694 8.73 29.59 16 5,891 10.96 40.55 17 7,639 14.21 54.76 18 6,888 12.82 67.58 19 6,920 12.88 80.45 20 4,269 7.94 88.40 21 2,687 5.00 93.40 22 1,498 2.79 96.18 23 894 1.66 97.85 24 506 0.94 98.79 25 310 0.58 99.37 26 170 0.32 99.68 27 86 0.16 99.84 28 40 0.07 99.92 29 24 0.04 99.96 30 11 0.02 99.98 31 6 0.01 99.99 32 4 0.01 100.00 Total 53,745 100.00 The smallest value of the index demonstrates the preference for egalitarian attitude, the largest - traditional. Because of the large number of characters in the index, it is poorly readable, which made us transform it into a binary variable. 4 step: Recoding grolesind by using the command “recode grolesind (8/20 = 1) (20/32 = 2)”, here 1 is for an egalitarian, 2 is for a traditional attitude to gender roles. I am not sure if this kind of reduction is not too much. You have most people “in the middle”, so neither egalitarian, nor traditionalist. At least three categories would be more appropriate, I personally would choose five. Table 2. Gender roles The same comments as for previous table grolesind Freq. Percent Cum. 1 47,509 88.40 88.40 2 6,236 11.60 100.00 Total 53,745 100.00 Name the categories! I really don’t want to decipher what 1 or 2 means. Use common language as often as possible J After all our transformations, we have the variable that shows “Attitude towards gender roles” which could be analyzed with the variable “Current legal marital status”. Correlation between these two variables was shown by crosstabulation. In our case we identify the variable “Current legal marital status” as independent, which is why we used row presents in our analysis. We also added column presents for comparison. No, no, no, please don’t J Choose one type of percentage and show only this one. If you add both percents into table and keep the frequencies, it’s really unreadable J 7 step Tubulattion of gender roles attitude index together with marital status by using analytical weight: tab grolesind v313 [aw=weight_g], row col Table 4. The maritial status and attitude to gender roles Grolesind Married Registere Widowed Divorced Separated Never maried Total 1 25,577.66 988.2956 4,367.487 3,187.015 680.39607 12,217.8 47,018.66 54.40 2.10 9.29 6.78 1.45 25.99 100.00 87.22 93.02 84.12 90.91 88.09 91.42 88.34 2 3,752.491 74.123351 824.19793 318.65666 91.981168 1,146.893 6,208.3431 60.44 1.19 13.28 5.13 1.48 18.47 100.00 12.79 6.98 15.88 9.09 11.91 8.58 11.66 Total 29,330.15 1,062.419 5,191.685 3,505.6714 772.37724 13,364.69 53,227 55.10 2.00 9.75 6.59 1.45 25.11 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 1, table format etc as in previous cases 2. name the values (1/2) 3, choose only row percentage and delete everything else, the column percentage are absolutely irrelevant and no one cares about frequencies Results As we can see from the table number 4, people that were never married have the highest percentage of the egalitarian attitudes(91 %) and the lowest percentage of the traditional ones. Then, it is demonstrated that when entering marriage, the percentage of people who stick to the egalitarian attitude lowers (87%). Taking this into account, we can state that people who contracted a marriage start holding onto more traditional attitudes towards gender roles than it was before that. On the other hand, when a person unmarries, he or she is prone to adopting more unconventional attitudes towards gender roles (91 %). Hence, the number of people who „think offbeat” increases in comparison to the one in marriage. What is interesting here is that death of a partner has a reverse effect; people, whose partner died, start to think more conventionally, which means that the percentage of people who have egalitarian attitude lowers (84 %). In conclusion, our hypotheses have been asserted. In addition to this, during our analysis we noticed other possible changes within these variables. Subsequently, we will do a research on these changes by means of multi-variable analysis. I am not sure about the causality direction, maybe people with traditional attitudes are more willing to get married? I am taking 3 points: 1 for table format, 1 for not recoding the index correctly (the huge middle part is clearly seen in your tabelation).1 for the results presentation (tables with too many information) +2 points bonus, you are the only one who used some recoding and made an index