The ability to read a scientific article and understand the statistical content is crucial. When one reads the literature, one does not only encounter familiar statistical procedures. Often one encounters unfamiliar procedures, and then one’s task is to understand as best as possible the link between the procedure and the research question, as well as the basics of how the statistical method works.

The goal of PST011 is to introduce various advanced statistical topics through reading scientific literature and discussion. It differs from PST010 in format; while PST010 was a lab-based module, PST011 is a discussion-based module.

Lecturers

Your lecturers are listed below. If necessary, you may contact them via email:

Lecturer Website Email
R. Morey (coordinator) [website]
M. Lewis [website]
M. Gruber [website]
B. Crawford [website]
C. Chambers [website]
I. Smillie [website]

Forum information

The PST011 forum on Learning Central will act as a central conduit for information and discussion. If you have questions about the course please post them there, so that others can benefit from the answers. The lecturers will monitor the forums and provide help.

Brief schedule of topics

The meetings will occur in-person in Tower 1.01.

  Date Time Lecturer Topics
1 10 Feb 10:00-12:00 M. Lewis Multiple regression
2 24 Feb 10:00-12:00 M. Gruber Triangulation in a quantitative context
3 10 Mar 10:00-12:00 B. Crawford To be announced
4 24 Mar 10:00-12:00 C. Chambers Transparency and reliability in science
5 28 Apr 10:00-12:00 I. Smillie Mixed methods research

Each meeting will consist of a short introductory lecture, a question and answer session, and a participatory discussion.

Preparation for each practical

It is expected that you will come to each practical prepared. Because a major component of the module is discussion, if you haven’t don’t the preparation you will not be able to participate. This will make for a somewhat awkward meeting, and it will make it harder for you to prepare for the exam. Discussing the topic is a great way of reinforcing knowledge, but if you haven’t done the reading you’ll have nothing for the discussion to reinforce.

Each meeting will have several readings of the lecturer’s choosing.

Any of the preparatory material may be examined.

Review after the practical

A “study question” for each practical is provided. In the days after the practical, try to answer this question in short essay form. Do your best to write a thorough, structured answer that shows that you grasped the material, and try to write it in 10 minutes, without access to your notes. The question is meant to make you think more deeply about the material and to help you prepare for the exam. The exam will be a series of short essays, so it is crucial that you practice writing about the course material in this format.

Helpful additional reading

In addition to the review, a number of additional readings may be given. These may be readings to help solidify your understanding of the week’s content, interesting readings to spur your thinking about problems in statistical analysis, or even games or interactive applets.

The content in the further reading will not be examined.

Meeting 1: Multiple regression

Preparation

Material type Content Link
Text reading Sections (including all subsections) 8.5,8.6, 9.9-9.11 of Field, A. (2017). Discovering Statistics Using IBM SPSS Statistics. 5th Edition.
Article reading Foo, Y. Z., Simmons, L. W., & Rhodes, G. (2017). The relationship between health and mating success in humans. Royal Society open science, 4(1), 160603. Royal Society Open Science
Video 1 Oxford Academic (2015). Bivariate and Partial Correlations. YouTube
Video 2 Quantitative specialists (2014. Multiple Regression in SPSS - P-Value (Part 3 of 3; You can also watch Parts 1 and 2 for context) YouTube
Video 3 HowToStats (2016). Multiple Regression Versus Multiple Correlation - Explained YouTube

Pre-session exercise: Are better looking men more successful in mating?

Download Foo et al.sav (also available from LearningCentral under “Handouts”) into SPSS.

Foo, Simmons and Rhodes (2017) wanted to explore whether health plays a role in sexual selection. If it does then more healthy people will be more attractive and have more mating success. To investigate this they obtained data for 101 men and 80 women. These data included:

  • Gender
  • Age (18 - 35 years);
  • Number of sexual partners (a measure of mating success);
  • Facial attractiveness (rated by people of the opposite sex);
  • Immune function (measured from a saliva sample);
  • Semen quality (for men only - obviously);
  • Frequency of Exercise (per week); and
  • Use of recreational drugs (yes = 1, no = 0: yes, you can do regression with categorical data).

Regression step 1

  1. Conduct a multiple linear regression on the male data looking at how number of partners M_Partners is predicted by age M_Age and attractiveness M_Attract. Does attractiveness predict reproductive success in this model?

Regression step 2

  1. Conduct another multiple linear regression as before but with additional predictors of health, semen quality, exercise and drug use: M_Immune, M_Semen, M_ExerFreq and M_Drug. Does attractiveness predict reproductive success in this model?

Overall, what should we conclude?

Review

If a multiple \(R^2\) is significant in a multiple regression, why might the individual predictors all be non-significant and what steps could be taken in such cases to further interrogate the data? Explain the risks of using stepwise regression in a psychological research setting.

Further reading

  • Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2013). Applied multiple regression/correlation analysis for the behavioral sciences. Routledge. Cardiff University Library

Meeting 2: Triangulation in a quantitative context

Preparation

Material type Content Link
Article reading 1 Please find and read a research article that studies ‘curiosity’. The article can be in any area of psychology, developmental psychology, neuroscience, personality, clinical psychology, computational modeling or even artificial intelligence related to psychological processes. It should address a question about understanding the effects of ‘curiosity’. Please bring the paper to class. You should be prepared to give a brief 1-minute summary (no slides necessary) to a smaller group during group discussions.
Article reading 2 Kidd, C., & Hayden, B. Y. (2015). The Psychology and Neuroscience of Curiosity. Neuron, 88(3), 449–460. Neuron
Article reading 3 Gruber, M. J., & Ranganath, C. (2019). How Curiosity Enhances Hippocampus-Dependent Memory: The Prediction, Appraisal, Curiosity, and Exploration (PACE) Framework. Trends in Cognitive Sciences, 23(12), 1014–1025. Trends in Cognitive Sciences

Review

Consider the article you chose to read and take to the meeting. You just need to give a brief informal summary of the article to the class (no slides or other visual aids needed). Consider how the article investigated ‘curiosity’ and think about potential advantages and weaknesses about the approach of the paper. Consider the general take-home messages from Article 2 & 3 and their approach to stimulate a new line of research.

Meeting 3: Topic to be announced

Preparation

To be announced

Review

To be announced

Further reading

To be announced

Meeting 4: Transparency and reliability in science

Material type Content Link
Article reading 1 Munafò, M. R., Nosek, B. A., Bishop, D. V. M., Button, K. S., Chambers, C. D., Percie du Sert, N., Simonsohn, U., Wagenmakers, E.-J., Ware, J. J., & Ioannidis, J. P. A. (2017). A manifesto for reproducible science. Nature Human Behaviour, 1(1), 1–9. Nature Human Behaviour
Article reading 2 Chambers, C. D., & Tzavella, L. (2022). The past, present and future of Registered Reports. Nature Human Behaviour, 6(1), 29–42. Nature Human Behaviour
Video preparation (47 minutes) “The System” (2008 TV special featuring Derren Brown) YouTube

Preparation: Consider an influential research article in your field (any article of your choice). Consider the following questions:

  • How do you decide how much you believe the authors’ claims?
  • What steps do you feel the authors could have taken to make their claims more trustworthy?

Bring this paper to the meeting ready to discuss your thoughts about these two questions.

Review

What so you see as the major pros and cons of Registered Reports as a tool for improving transparency and reliability? Explain one pro and one con.

Further reading

Meeting 5

Material type Content Link
Video What is Mixed Methods Research? (15min video with John Cresswell) YouTube
Text reading Ch. 4 from Plano Clark, V.L. & Ivankova, N.V. (2016). Mixed Methods Research: A guide to the field. Sage, London. SAGE Research Methods
Article reading 1 Guerra, N. G., Williams, K. R. and Sadek, S. (2011). Understanding bullying and victimization during Childhood and Adolescence: A Mixed Methods Study. Child Development, 82: 295-310. Wiley
Article reading 2 Article of your choice (see below)

Preparation

Identify a mixed methods paper of your choice (ideally from your own field); use search terms like mixed methods or qualitative and quantitative (plus terms relating to your field, e.g., developmental, cognitive). Please bring this paper to the meeting. You should be prepared to give a brief 1-2 minute summary of your paper (no slides or visuals required) to a smaller group during group discussions.

Review

Consider both article 1 and your chosen article 2. Generally, what do you consider as the advantages of combining qualitative and quantitative methods in a mixed methods design? What do you see as the potential advantages and possible limitations of mixed methods research designs in your area of study?

Assessment

The final module mark for the module will be based on 30% coursework (essay) and 70% exam.

Essay

Find a scientific article that interests you (either a published one or a preprint) that uses one of the designs or techniques discussed in PST010 or PST011. Identify the primary scientific question of interest and the primary statistical test (or method, if it isn’t statistical).

Due date

The due date for the coursework will be in Week 7.

Essay organization

The essay will be a critique of the connection between the primary scientific question of interest and the statistical test/method that was used to illuminate that question. The essay should be about 600 words.

Organize your essay in the following way:

  • Introduction: Describe the primary research question in the article, and briefly describe the statistical method used to illuminate or inform about that question.
  • Critique the link between the scientific question and the statistical test: How strong is the link between the scientific question outlined in the paper and the method? Are there issues that threaten the scientific conclusions (e.g. confounds)? Is the statistical test too narrow to draw useful conclusions (e.g., a test of population means where a more interesting test should have been performed)? Or do the researchers deserve praise for a particularly clever test of their question or sophisticated method?
  • Further work: How might you build on the work of the researchers to either provide a better test of the scientific question, or to apply their methodological insights to a new domain?

Essay Marking

Definition of key terms:

  • Knowledge and understanding: Using knowledge of relevant theories and literature, demonstrating an understanding of the relevance and success of theories in the broader dimensions of the area, evidence of understanding the limits of current understanding.

  • Analysis and evaluation: Demonstrating awareness of gaps or limits in the knowledge base, selecting appropriate methods of enquiry, presenting a lucid rational argument with clear well thought out conclusion and an appreciation of the future direction of work in the topic area.

  • Synthesis: Demonstrating the combination of different viewpoint/levels of analysis when dealing with complex and perhaps conflicting information. Considering and identifying the appropriateness of methods and/or experimental design.

  • Originality and innovation: The demonstration of original thinking – that is, evidence of independent analysis of information, and presenting new ideas and or the application of techniques in novel ways to address specific problems.

  • Independent learning: The piece of work demonstrates evidence of relevant reading and reflection on material outside the lecture content or essential reading.

Marking scale

A mark of 50 or above is considered “passing”.

Distinction (70-100): An essay marked with distinction should, at a minimum, be correctly formatted throughout with few minor errors. The rationale for the study and the statistical analysis should be developed accurately and succinctly. The work should show (1) evidence of accurate knowledge and understanding of the key statistical concepts. There should be (2) evidence of synthesis of material and (3) critical analysis (e.g., the strengths and limitations of various approaches one could take data analysis) of relevant material that is justified and that connects to the research question of the chosen paper. The work should demonstrate (4) independent learning and synthesis of material (5) and provide a logical discussion regarding the limits of the statistical approaches and implications that go beyond practical handouts/lectures or core texts.

Higher marks with distinction (80 and above) should display sustained quality in all of the areas described for distinction with no errors or critical omissions. Relevant concepts should be discussed and analysed in a critical fashion, e.g., competing/alternative analyses should be presented clearly and cogently discussed. There should be evidence of independent learning accompanied by original insight into problems and solutions. The consideration of potential alternative approaches should clearly display evidence of originality in terms of identifying problems and/or proposals for novel analyses.

Very high marks with distinction (90-100) should show all the characteristics of discussed above and, in addition, present clear evidence of sustained innovation and originality in terms of knowledge, understanding (e.g., linking concepts in novel ways) and critical analysis (insight into issues or the appropriateness of methods and analysis) that is clearly based upon independent learning and demonstrates novel synthesis of information. Consideration of potential alternative approaches should clearly demonstrate sustained originality and innovation.

Merit (60-69): A essay in the merit category will, at a minimum (60), display the majority of the following characteristics: (1) It should have only minor omissions or errors; (2) the content should be appropriate, accurate and convey the research question and statistical content clearly but may lack depth. The discussion will be largely appropriate but may need greater depth and scope. Consideration of the implication of the arguments will be limited and will reflect information provided in lecture content/hand out material. A better answer within this class should be more comprehensive, show evidence of depth of understanding of core material and (4) clear evaluation in developing the rationale for the report (e.g., contrasts are made between approaches). (5) Good answers within this class should develop its topic clearly and illustrate the above together with evidence for some independent learning and synthesis of relevant material. The general discussion should reflect on the rationale for the critiques and connection with broader scientific practice. The discussion should demonstrate critical understanding of lecture/core texts and recommended readings.

Pass (50-59): A minimum pass (50) should (1) be clearly and succinctly written and appropriately formatted but will have some errors and omissions. The essay should (2) show evidence of a basic understanding and application of relevant material to explain the statistical analyses. The level of knowledge (e.g., reference to key statistical concepts) shown in the introduction and discussion is limited in scope and depth, and may be incomplete and does not go beyond lecture material/core texts. Better answers will show evidence of synthesis but it will not typically extend beyond that provided in lectures/core texts. There may be some irrelevant arguments, and may contain factual errors and/or omissions. Higher marks within this category should demonstrate evidence of (1) an accurate understanding of the material and its appropriate application to the statistical report (e.g., demonstrate the knowledge of key concepts and their underpinnings), (2) but should lack depth and critical analysis and contain some errors or omissions. Any consideration of the broader implications of the arguments may be weak or not fully developed.

Failing (40-49): A failure with high mark (40-50) should be formatted correctly but with substantial omissions or inaccuracies. The introduction will show some but limited knowledge of material directly related module material. The development of the rationale for the study and analysis should be weak (there is no reflection on basic statistical concepts and their underpinnings) and the material should be limited to lecture/handout content. There should be no evidence for independent learning.

30-39: A piece of work that contains extensive errors and omissions and fails to address the question or topic. There is only a small amount of acceptable or relevant information. There is no evidence of appropriate critical awareness or analysis and scant use of lecture material. The answer will be poorly structured with minimal integration of relevant ideas and little or no evidence of understanding the wider context. The work may demonstrate a failure to follow instructions appropriately.

20-30: A piece of work at this level should fail to demonstrate sufficient relevant knowledge and understanding to address the question, and have serious errors and omissions. There may be evidence of misconceptions, errors and irrelevancies.

10-20: A piece of work that is based at this level will reflect minimal appropriate knowledge beyond that one might expect from a lay person. There is little, if any, appropriate argument and the answer does not address the question. There is no evidence of conceptual understanding. Any relevant knowledge is marred by serious errors.

1-9: A piece of work that displays only vague application of relevant knowledge. Any relevant knowledge should be compromised by serious errors and omissions and/or a failure to follow instructions. There are few if any relevant ideas and those present are vague and expressed with no evidence of critical awareness or the ability to link ideas.

0: There is no answer. There is no evidence of relevant knowledge and understanding for psychology, or does not answer the question set.

Essay feedback

Generic feedback will be given on essay marks; however, any student wishing to discuss their mark with one of the instructors should make an appointment via email.

Examination

The exam will consist of approximately ten short essay questions. These questions will cover the content of the review lectures at the beginning of each meeting as well as any preparatory reading. You will be expected to write several paragraphs showing your understanding of the material in the preparatory readings and practicals. For general marking guidelines, please see your handbook.

This document was compiled Wed Jan 26 10:02:40 2022 (Europe/London).