Design your Own MLR & Dummy Coding
- Conduct a meaningful MLR analysis (of your own design)
and be able to interpret the analysis - explain it to someone next to
you or your tutor.
- For the above analysis, obtain and interpret an analysis of
multivariate outliers (use Mahalanobois and Cook's Distance - see
notes in previous tutorial)
- Conduct a meaningful MLR analysis which requires you to recode one
or more of the variables as dummy variables.
- For either of the above analyses, conduct the analysis using
each of Direct, Forward, Backward, Stepwise, and Hierarchical methods.
Be able to interpret and analyse why or why there aren't differences
between the output for each method.
- Work out how to produce stand-alone partial correlation tables in
- Produce stand-alone correlations and partial correlations from any
dataset and be able to depict this data in a Venn Diagram.
Critical Reading of MLR in Articles
- Find a published research article which has used Multiple Linear
- Briefly summarise, explain and critique the articles' use of
Multiple Linear Regression. You should include and
interpret the articles' APA reference details, describe variables,
summarise the research question/hypothesis, relevant statistics and relations
amongst the variables. Also emphasize what the results mean.
You are also encouraged to make critical comment, e.g., Is this use of
regression appropriate? Could/should the analysis have been
conducted differently? (these questions are only suggestions)
- IMPORTANT: You will be asked this question in the MLR2 quiz - it
will be worth 1.5 marks. You may pre-prepare your answer and
then cut and paste from an electronic attachment during the exam
(e.g., beforehand email it to yourself, bring on a floppy disk or data
key). Alternatively you could print it out and then retype in
during the exam.
- Word limit: 250 words - longer answers will have a penalty