Exam Prep

Data Analysis for Psychology in R 3

Josiah King & John Martindale

Psychology, PPLS

University of Edinburgh

Course Overview

multilevel modelling
working with group structured data		 regression refresher
introducing multilevel models
more complex groupings
centering, assumptions, and diagnostics
recap
factor analysis
working with multi-item measures		 what is a psychometric test?
using composite scores to simplify data (PCA)
uncovering underlying constructs (EFA)
more EFA
recap & exam prep

Outline

  • Process

  • Structure of exam

  • Things to take into exam

  • Exam “strategy”

  • Example Questions

Process

Details

WHEN?

Date: Friday, 20th December 2024 Time: 1:00 p.m. to 3:00 p.m.

WHERE?

Lennox, Lammermuir & Moffat Room, Edinburgh International Conference Centre




CHECK YOUR CALENDAR https://www.ed.ac.uk/timetabling-examinations/exams/exam-diets

Details

  • Arrive in plenty of time

  • There will be direction to room

  • When you go into the room, you will need your student ID on your desk

  • Invigilators will give you exam instructions

Things to take to exam

  • Student Card!!

  • Pens (multiple) blue or black ink

  • Pencil (for rough work)

  • Ruler

    • If you like, not really needed, but might be useful.

  • Calculator

    • Any type (see list on Learn of permitted calculators)
    • No advantage to scientific vs normal
    • Phone/Watch can not be your calculator

Things you will be given

What you are given

  • a specific sheet for your MCQ

  • an answer book for all other questions

  • a rough work book ( if you want to use it )

  • the equation sheet

FILLING IN FRONT COVERS

MCQ sheet:

  • matriculation number (UUN but without the “s”)

Answer book:

Things you can take away



NOTHING!

(Other than the things you bring in with you like pens, pencils, calculator)



When you finish, you leave all papers (questions, answers, equation sheet) in the exam hall.

Questions?

Exam Structure

Exam Structure

Section A - 30 MARKS

  • 15 MCQ

  • Just like quiz questions on Learn

  • On all topics

Section B - 70 MARKS

  • 8 Questions

  • Marks range from 4 to 15 marks per question

Exam Strategy: Section A

Section A - 30 MARKS

  • 15 MCQ

  • Just like quiz questions on Learn

  • On all topics

15-18 minutes

  • You can answer these quickly at the end
  • Do the ones you know, move on
  • Keep track of how many you have not done, so you can go back to them at the end

Exam Strategy: Section B

Questions on:

  • Some short calculation
  • Some ask you to explain statistical concepts discussed in the course
  • Some give you a research design/question and task you with specifying an appropriate model
  • Some give you model output and ask you to provide interpretation

Section B - 70 MARKS

  • 8 Questions

  • Marks range from 4 to 15 marks per question

content questions
Weeks 1-5 (multilevel models) 4 questions totaling 32 marks
Weeks 7-11 (pca, efa, psychometrics) 4 questions totalling 38 marks

Exam Strategy: Section B

Questions on:

  • Some short calculation
  • Some ask you to explain statistical concepts discussed in the course
  • Some give you a research design/question and task you with specifying an appropriate model
  • Some give you model output and ask you to provide interpretation

approx 5 marks

show your working (could get marks for this if answer is wrong)

think, e.g.:

  • calculating icc from a intercept only model
  • calculating coef from fixef and ranef
  • calculating variance explained from eigenvalues
  • calculating SS loadings

Exam Strategy: Section B

Questions on:

  • Some short calculation
  • Some ask you to explain statistical concepts discussed in the course
  • Some give you a research design/question and task you with specifying an appropriate model
  • Some give you model output and ask you to provide interpretation

approx 6-8 marks

often these ask to define a couple of things.
1 to 2 sentences each thing.

think, e.g.:

  • “ICC”
  • “random effects and fixed effects”
  • “no pooling/complete pooling/partial pooling”
  • “factor loading”
  • “reliability”

Exam Strategy: Section B

Questions on:

  • Some short calculation
  • Some ask you to explain statistical concepts discussed in the course
  • Some give you a research design/question and task you with specifying an appropriate model
  • Some give you model output and ask you to provide interpretation

approx 10 marks

you have done this in the report, and lots of times in labs.

Hint: PCA/EFA are not “research design/question”-based in the sense of estimation, so these questions are likely to be multilevel models.

You probably know more than you think you do.
- e.g. start with an outcome variable.
- ask yourself: “what in my question is asking about ‘effects on Y’/‘influences on Y’/‘predicts Y’”.

Exam Strategy: Section B

Questions on:

  • Some short calculation
  • Some ask you to explain statistical concepts discussed in the course
  • Some give you a research design/question and task you with specifying an appropriate model
  • Some give you model output and ask you to provide interpretation

approx 10-15 marks

you have done this in the report, and also seen it plenty of times in the labs and lectures

  • the questions will list the things we expect comments on
  • don’t just parrot back numbers. provide interpretation
  • these usually provide information about the research aims, so place your answer in that context.

These are going to be our main functions we have seen throughout DAPR3: lmer(), principal(), fa(). Print an output from the lab/lectures and annotate it. what does each bit of the output represent?

Questions?

Mock Questions