4. Carry out a planned standard factorial contrasts analysis, controlling the familywise error rate at α = .05.

•   Analyse your data using the PSY statistical program. Base your analysis on tests and raw confidence intervals.

•    In order to make your confidence interval inference more meaningful you should

propose a minimum difference in dependent variable units of theoretical or practical importance.

• Your analysis must produce at least one significant main effect contrast and at least

one significant interaction contrast. What happens to the other contrasts is up to you. Whatever the outcome, the results should be plausible.

•    It is up to you as to whether or not each confidence interval corresponding to a significant contrast establishes a theoretical/practical important effect.

NOTE 1: You may find that you need to modify the data in order to achieve the results you are looking for. A very large F ratio for an individual contrast would usually indicate that MSE (mean square error) is implausibly small, given the pattern of differences between means. For the purpose of this exercise, an F of 100 or greater will be regarded as implausible.

NOTE 2: As you are asked to make CI inferences for your contrasts, this precludes including trend contrasts in your analysis.

Write up of the Exercise (this is what you submit to Turnitin as a single document)

You should submit to Turnitin a single document that includes all of the following (in this order):

i. Your name and student ID number

ii. A brief description and rationale for the design, clearly stating the independent and dependent variables and how these are to be operationalised.

iii. An account of the planned analysis, including a rationale for the planned contrasts chosen for the analysis (i.e., why have you included these contrasts, in particular?)

iv. A clearly labelled table of A, B and AB contrast coefficients, identifying which contrasts are included in your analysis (and make clear which levels of Factor A and Factor B each of the JK coefficients refers to).

v. A summary of the data (including a clearly labelled table of means and plot of cell means).

vi. A  summary of the outcome of your contrast analysis (both tests and raw confidence intervals). State the decision rule used for tests. Include the PSY summary  table  and formatted PSY confidence interval table (edited from PSY output to produce a single table for A, B and AB contrasts).

Provide a concise interpretation for your contrast analysis (what does it all mean?). Your interpretation should include directional and confidence interval inference.

vii. Appendix – include a copy of your PSY input file(s), and complete (unedited) PSY output file(s) at the end of your assignment (as an Appendix).  Include  PSY output for each  run through PSY. Do not delete any part of the PSY output. Hint – paste PSY output into your Word doc as Courier 10pt font. This will make sure the output has the correct spacing.

NOTE: Your Assignment cannot be marked unless your PSY analyses can be verified.

Make sure all your tables and graphs are clearly labelled.

Include in your report:

.   table of contrast coefficients

.   a table of cell means

.    plot of cell means

.   contrasts Summary table (as produced by PSY)

.    Raw CIs table (edited from PSY output to produce a single table for A, B and AB contrasts)

IMPORTANT – It is your responsibility to make sure you upload the correct document to the Turnitin link (with the correct PSY input and output files included in the Appendix).

Note that no resubmission will be allowed after the deadline.

Length: The written part of your report (excluding tables and graphs and Appendix) should be no more than about eight typed pages (1½ spacing).

Evaluation: Reports will be evaluated on the basis of:

.    Rationale of the design. (Do your two factors make sense in combination with each other? Are the levels of each factor sensible and appropriate to your research question?)

.   Clear description of variables. (How is the DV measured? How are the factors and factor levels operationalised?)

.    Rationale for your planned contrasts.

.    Plausibility of the data (Are the cell means and within-cells variability plausible, given the nature of the DV and factors?)

.   Correct execution of the planned contrasts analysis. Correct decision rule for tests? Have CIs been correctly generated?

.   Clarity and conciseness of the interpretation of the contrasts analysis.

Rambling or lengthy reports will not be highly regarded. You do not need to include a long introduction or a literature review. If you wish to refer to one or two key references in the course of showing the relevance of the design to (say) a theory, then by all means do so, but in most cases this should not be necessary. Provide enough background information to enable the reader to understand the logic of the design and how the variables were operationalised, but no more than this.

How to Avoid Plagiarism and Student Misconduct

Do not use designs and data sets referred to in the PSYC3001 course materials (if you do, you may receive a mark of 0).

The data for this Exercise should come from you (no other source). However, for inspiration in choosing variables, you may refer to other Psychology courses or the psychology literature.

You are reminded that the work you submit must be your own and should not be the work of another student (past or present) or any other person.