Research Methodology
Introduction to Research Methodology Assignment II – Analysis In assignment I your skills in research design were tested, Assignment II tests your skills as statistical analyst and reporter. On the basis of an existing dataset (on the 2009 European Parliament elections) you will show your abilities to analyse data and report on the findings in a meaningful manner. European Elections Survey This survey was administered shortly after the 2009 European Parliamentary elections (4 to 7 June 2009). Respondents were asked all sorts of things related to Europe, the elections, but also their media use and other personal characteristics. Thus, the dataset consists of a large number of variables that can either serve as dependent or independent variables. The data you will be analyzing will not be on all countries of the European Union, but only on one very special country: Great Britain. That’s right, as a proper farewell, you can have a look at the Brits’ views (almost) 10 years ago, and see whether they still liked us then (or not). Goal of the assignment Your task is to employ the statistical techniques you’ve learned in this course in a meaningful way. For each of the techniques listed below, present: What you will analyse An overview of the variables you will use Your results Your interpretation of the results You are not restricted to one overarching research question: for each technique you can think of a new question you will answer. You will be tested on whether you are able to apply the different statistical techniques correctly, showing properties or relationships that make sense, select the proper variables for each technique, and whether you can report and interpret the statistical findings meaningfully. The techniques you must use are: a) univariate descriptive statistics (mean, median, mode) b) cross-tabulation, including controlling for a third variable use of proper measure of association in a cross-tabulation c) t-test, dependent & independent samples d) analysis of variance (ANOVA) one-way & two-(or more) way e) multiple regression (including at least three independent variables) f) Factor analysis (using at least four items) and reliability analysis g) In addition: recoding, computing and selecting cases, where applicable Sections As mentioned, we do NOT expect you to use all of these techniques in a single coherent research effort: It is perfectly acceptable if you segment your report in several sections, in which you discuss what you will examine in that section (e.g., “association between this and that, controlling for this ...”), followed by a report of the findings. The separate sections need not be related, and you may combine several techniques per section (you couldforinstancereportandinterpret descriptivestatisticsforthevariablesusedinacrosstab,oraregression). Please take note of the following: Always first describe what you will analyse (this can be very brief: I will assess the influence of education and income on frequency of newspaper readership, controlling for sex, using a regression analysis with variables xxx, and...), then report the findings in text and tables, and include your interpretation of the findings. You do not need to give a theoretical justification for the analysis, but it should make sense. E.g., explaining income using education and sex is fine, but sex as a dependent variable is not. Explaining educational level by somebody’s exposure to a particular newspaper does not make sense either, while the reverse may make more sense. Tables you present in your report should only contain what is required (check the Reporting Results guidelines), and should be made up in Word or whatever text editor you use. NEVER include ‘raw’ SPSS output in your report. Include an appendix with the SPSS SYNTAX used for your analyses. Include syntax statements for all your analysis Include comments in between the syntax commands to indicate for which section of your paper the syntax was used Make sure you also include recode or compute statements for variables you changed or created Also indicate the missing values you declared – either though the syntax, or by including a comment: E.g.: * For variables Opinion1 & Opinion2 the values 8, 9, 10 are declared missing. In your report, make the tables pleasant for the reader to go over, only include the information required and always include informative headers, not SPSS terms or variable abbreviations (did I mention you should NOT cut & paste SPSS output in the assignment?) The length of your report (excluding the SPSS appendix) will depend on the variables you use and the size of your tables, but somewhere around 10 pages would be a reasonable length. For instructions of what to include when reporting your analysis, refer to the Reporting Results guidelines (see blackboard for the link, the lecture slides and the book (Field). Please note: this is an individual assignment. You are not allowed to hand in the same assignment or sections as another student. If you copy the work, words or research questions from another student, you will fail this assignment. Your analyses, report and interpretation should be your individual work. There are many options for you to explore on the basis of this dataset. Many separate variables are suited for analysis as is, or inclusion into a composite measure or scale. Feel encouraged to make creative (yet sensible) use of the data.