|
All the requirements and instructions are shown in the document. I've already done the calculating stuff and u need to interpret them follow the instructions.
ADVANCED QUANTITATIVE METHODS
MINI PROJECT EXERCISE
DO NOT use raw PASW output in your answers. You should aim to write up your work in the style of a short research report. References to the social policy literature are not necessary, as the main purpose is to test your skills in using and understanding quantitative methods, but would be nice "icing on the cake" if you could dig out some relevant literature on housework and gender roles and may help you get an even better grade.
This exercise asks you to imagine that the Department for Business, Innovation and Skills have asked you to produce a short research report on the factors influencing the amount of time people in England spent on housework in 1999. The Department is particularly interested in the relationship between University education and housework, and officials at the Department’s regional office in Yorkshire are particularly interested in whether people did more housework in Yorkshire compared with the rest of the country.
Your over-arching research question is:
• What factors influenced how much time people in England spent on housework in 1999?
The Department would also like you to address the following specific research questions:
Question A – Bivariate Hypothesis Tests
1. Was the amount of time people spent on housework each week significantly associated with living in Yorkshire (which includes everyone living in South Yorkshire, West Yorkshire and the Rest of Yorkshire and Humberside)?
(ii) Was the amount of time people spent on housework each week significantly associated with equivalent household income after housing costs?
Question B – Factorial ANOVA
(i) Was there an association between time spent on housework and having a university degree?
(ii) Was this association significantly different for men and women?
Question C – Multiple Linear Regression
(i) Did the apparent “Yorkshire effect” on housework remain significant after controlling for equivalent household income and other individual characteristics?
(ii) What conclusions can you draw from your multiple regression analysis about how different factors influenced time spent on housework in 1999?
You should write up in the form of a short research report, using the following headings and sub-headings: 1. Data, 2. Methods, 2.A Bivariate Hypothesis Tests, 2.B Factorial ANOVA, 2.C Multiple Linear Regression, 3. Results, 3.A Bivariate Hypothesis Tests, 3.B Factorial ANOVA, 3.C Multiple Linear Regression, 4. Discussion.
“Methods” section should state, within each relevant sub-section, what statistical techniques you use, why they are used, and how they are used. It should justify the use of these techniques in relation to their key assumptions. For each BHPS source variable included in your model, you should (a) describe what it measures or indicates, (b) describe any important variable transformations (e.g. recoding a multinomial variable like highest academic qualification into a binary variable indicating whether nor not the person has a university degree, e.g. computing gender interaction terms), and (c) explain your rationale for choosing to include it in the model i.e. #p#分页标题#e#指导Assignmentwhy it might be expected to influence hours of housework.
Keep your “Results” section brief: select the key important results to present, making judicious use of tables and charts, and try to stick to the bare undeniable facts about what you found (e.g. associations and their significance). Leave the discussion of those facts to the “Discussion” section – i.e. discussion of what your results can and cannot tell us about the real world.
Make sure all tables are clearly labelled and carefully crafted. Use real words, not computer variable names. Above all, do NOT include unedited PASW output in your report. Re-format tables yourself, either using Excel or PASW table editing features. You will lose marks if the examiner suspects you have cut and pasted irrelevant raw PASW output that you do not fully understand. You do not need to present tables for bivariate hypothesis tests: these can be written up in sentence form. However, you must present factorial ANOVA and regression results in tables.
In the “Results” section, you should describe in words what your regression model tells you about the association between each explanatory variable (e.g. Yorkshire residence, income, education, etc.) and the amount of time spent on housework.
In the “Discussion” section you should consider the meaning and plausibility of the multiple regression results in a real social policy context, and come up with some intelligent guesses as to the nature of the possible causal pathways that may (or may not) lie behind the results. You should also consider the limitations of your model. How much “unexplained” variance remains? What other factors, not represented in your model, might also influence the amount of time spent on housework? How might the omission of these missing factors from your model potentially bias or invalidate your findings?
|
