Hello, dear friend, you can consult us at any time if you have any questions, add  WeChat:  zz-x2580

PSYC40005 Advanced Design and Data Analysis

Assignment 2
Preliminary steps:

1. Download and unzip “Assignment-2-R.zip” from Canvas.

2. Inside the “Assignment-2” folder, open the R Project file “Assignment-2.Rproj”
to start R Studio. Your working directory will be automatically set to this folder.

3. In R Studio, open “A2-Setup.Rmd”. Follow the instructions carefully.

4. You can then use “A2-Working.Rmd” as a space to write the code for your
assignment in and save it as you go. Please note that you should not submit this
R Markdown as your assignment; see submission details at the end of these
instructions.


The project, and criteria for marking:

The variables are named q1 to q30, and are self-rating items of behavioural tendencies
and preferences.

• PART 1 – 33% of marks: Conduct an exploratory factor analysis, reach a
preferred model, and interpret the results.
• PART 2 – 45% of marks: Using the EFA you did in Part 1 as a guide, run a
confirmatory factor analysis on a parsimonious factor structure for the data.
Describe the fit of the model and show the diagram. After testing the fit of your
original model, you should discuss what the outcome would be of engaging in
one round of model re-specification and re-fitting, but for the purposes of this
assignment you should not engage in multiple rounds of model re-specification
and re-fitting.
• PART 3 – 16% of marks (maximum of 250 words for this part): Reflect on what
you’ve just done. To what extent do you think your model is ‘confirmed’? To
what extent do you think your model is ‘overfit’? Do you think your findings
would replicate if we took a new sample? Why/why not?
• PART 4 – 6% of marks (maximum of 90 words for this part): Imagine you show
your model from Part 2 to a friend. Your friend creates a new model and tests it
on the same data. Her model has more parameters than yours and has zero
degrees of freedom and a Chi-Square of zero. Your friend says to you “My
model fits better than yours! Therefore, my model is more useful than yours”.
Write a brief reply to explain to your friend why she is wrong. In your reply,
include some reference to your own model from Part 2.


2/3
Within each Part a small number of marks are devoted to “Writing the report in
appropriate style”, and this amounts to 10% of marks in total. Style includes formatting
as well as writing style, spelling, grammar, diagram formatting, APA style formatting,
and so on.

The core content for the assignment is covered in Lecture 4 Exploratory Factor
Analysis, Lecture 5 Confirmatory Factor Analysis, and (to a lesser extent) Lecture 6
Path Analysis. You should also consider the material from the associated Lab Classes.
You can get a high H1 on the assignment without going beyond the course materials.
However, in relation to some issues a small number of bonus marks are on offer for
appropriately going above and beyond the lecture (e.g. following up on a reading
mentioned in the lecture, citing some relevant point that goes beyond what was covered
in the lecture). This is mostly relevant in relation to the issues of fit statistics and sample
size.

Within each Part, 90% of the marks are devoted to your coverage of various issues
raised in the aforementioned Lectures and their associated Lab Classes. Issues that
should definitely be considered in at least one Part include: introducing the reader to
your data, factorability of your data, rotation, selection of the number of factors, sample
size, significance testing, distributional assumptions, identification, methods of
estimation, assessment of global fit, assessment of local fit, bootstrapping, results
tabulation, model specification and re-specification, interpretation of results. You don’t
need to consider the issues in that order. You may consider other issues too, and the list
above should not be regarded as a complete ‘checklist’ of appropriate content.

Questions about the assignment can be made on the discussion board, and we will
maintain a thread called “README ASSIGNMENT 2” which will contain a list of any
posts we consider likely to prove useful for the assignment. We will also maintain an
errata section in the Canvas module for Assignment 2.