Assignment 3a (Lecture themes 9 and 10): Quantitative data analysis

Home Assignments for Module 3

· Based on the material of module 3, you will need to do one home assignment (number 3).

· For assignment 3, you can choose either one of the two available options (3a OR 3b).

· Deadline for the assignment: March 15.

· Submit the assignments by e-mail to your respective seminar leader.

· Please put in the e-mail subject field “homework 3 [your last name]”. Use the same format for naming your files (e.g., “homework 3 Ivanov.doc” or “homework 3 Johnson.zip”).

· For homeworks that contain more than one file (e.g., paper review + paper PDF or analysis description + SPSS output) it is recommended to use the ZIP format.

Assignment 3a (Lecture themes 9 and 10): Quantitative data analysis

Your task is to choose and apply a set of multivariate quantitative methods and describe your variables and findings as you would do in an extended research report. Allow at least 8 hours to complete this assignment.

Dataset. You can use the same dataset that you used for assignment 2 (as long as the data are appropriate for multivariate methods) or find a new one. Explore the dataset to choose variables which are 1) interesting for you, and 2) will allow you to use the data analysis methods described below. You will need to do all steps of the analysis described. For the steps d and e you have a choice of one of the methods suggested.

Methods. Perform the steps of analyses that you normally would when you do a research project:

a) Describe your sample and the set of variables;

b) Clean the data if needed, describe missing data (if any) and choose a strategy to deal with it;

c) Provide relevant descriptive statistics for your variables;

d) Apply and interpret one model predicting an interval variable using a mixture of nominal and ordinal/interval variables. You can choose one of the two options. For bonus score, you can try both and compare results.

a. hierarchical multiple regression with moderation analysis (for survey data);

b. Factorial ANCOVA/other type of General linear model (for experimental data);

e) Use exploratory factor analysis (EFA) to investigate the structure of a set of variables. Specify your theoretical hypothesis of the number of dimensions (if any) and choice of rotation strategy. Describe criteria you used to set the number of factors. Interpret and compare 2 models with different number of factors. For bonus score: use hierarchical cluster analysis (suggested approach for any data that can be used in EFA: Ward’s method, Squared Euclidean metric, data standardized to z score by variable) and compare the dendrogram with the results of EFA.

The product of your work will contain two parts:

1) research report: a description of your analysis and results (similar to what you find in the "Results" section of a research paper, but slightly more detailed), where you describe your analysis:

• formulate your research task(s) related to multivariate methods, e.g.: psychometric investigation of the structure of a set of items, or investigation of a set of predictors
• describe briefly the data (samples, variables) you will be using,
• describe the steps of your data analysis in a clear and replicable way,
• present the results in a textual, tabular, and/or graphic form,
• give a brief description of the results,
• formulate a short conclusion about the evidence provided by the data with respect to your research question(s).

When you describe your results, aim to present them in the way that you would use in a research paper (e.g., use APA formats for tables, rather than copy them from SPSS; try to provide and interpret the sizes of effects you find, etc.). Use APA structure for your text. The approximate length of results description is 2,000-4,000 words.

2) an appendix: please save and attach to your submission the SPSS outputs (.spo or .spv) containing all the steps of your analysis.

Date: 2016-04-22; view: 439