PSYC 422 Lecture 1

A/Prof. Deborah Apthorp

Slides are open source

Link for the HTML version of these slides (always up to date)

Hi everyone. Welcome to Psych 422 for 2024. This is the introductory lecture, lecture number one, and I’m going to present it in three separate parts.

Oh, just a note before I begin. These slides are actually open source, they’re written in R Markdown and Quarto. And if you click on this link or if you use the QR code, you can get to the html version of the slides, which is always up to date because I upload them to GitHub whenever I change anything.

So you can follow along and click through each slide as you follow the lecture if you’d like to do it that way. But I will also provide a pdf version for those who like to print things out. So I hope you enjoy that.

Outline

• 1.1: Things you need to know about PSYC422 (an incomplete list)
• 1.2: Psychology as a science - what characteristics of science are important?
  • How is Psychology doing on those?
  • Pseudoscience - how to spot it
• 1.3: Practical stats section:
  • Resources
  • Intro to jamovi and brief demo on my own data

Okay, so a quick outline of what I’m gonna talk about today. Firstly I’m just gonna tell you the things you need to know for this year about Psych 422.

This isn’t a complete list, of course. Pretty much everything you need to know is on the MyLearn site anyway, so you should be able to read through that and find it everything you might need to know, or you can ask me on MyLearn (I’m probably going to keep calling it Moodle) or you can email me. But that should be hopefully a last resort if you haven’t found your answer.

So the second thing I’m going to talk about is the first set of things that are in your online textbook. I’m going to talk about psychology as a science, and what characteristics of science are important, and how psychology’s doing on those. Also how to spot pseudoscience. The first topic for your Discussion Board covers this, and that’s usually a great fun topic that people get into.

And the third part of the lecture is going to be a practical stats section. I’m going to introduce you to some of the resources that you can use; and lastly I’ll give a quick introduction to jamovi, which is a statistical software that I’m going to encourage you to use this year, and I’ll give a brief demo of that with my own data.

Things you need to know about PYSC422

First I want to go through the things that you’ll need to know this year about Psych 422.

Assessments

• Online Quiz component - 20% (3 knowledge (m/c and short answer), 3 stats, mark = best 5).
• Discussion Board posts - 10% (about 1500 words, responses to prompts through trimester, min = 5 topics)
• Written assignment - critical thinking blog post - 30% (scaffolding workshop at Intensive)
• Final Exam - 40% (multiple choice and short answer, online)

The things everyone wants to know, of course, the assessments. So what are they?

There’s an online quiz component that’s worth 20% of your mark.

So there are three general knowledge type quizzes that come from lectures, textbook and reading content. These are mostly multiple choice with some short answer questions. I hope you’ll find them relatively straightforward and fun.

And three of them are stats exercises, so they’re also in quiz format, but you’ll get the stats material beforehand, you’ll be able to work through the problems and hopefully get your answers. These are untimed. The stats ones are untimed. You’ll have as long as you’d like to complete those. Obviously they’re due at a particular time which is on MyLearn; but you’ll be marked on your best five of those six quizzes.

Mostly people do pretty well on those, so I don’t want you to fret about them. They’re more an exercise to make sure you keep testing your knowledge.

Then 10% is the discussion board post. Now, these are actually my favourite part of this whole unit. You need to write about 1500 words in total for a minimum. But most people tend to write more than that. And each week I’ll put up a prompt for people to write about, you don’t have to do it within that week, but sooner is better than later. And I’d like you to also respond to other people’s posts in that forum as well.

You don’t need to write on every single topic, but you need to write on at least five. There’s a rubric as well, so you can look at that and make sure you are fulfilling the criteria on that. It doesn’t need to be very formal. In fact, I prefer informal discussion. Just get into it - get involved - and I think you’ll find that it makes you become part of the community of 422.

I’ll get in there and respond to posts as well, and I really enjoy reading them, so I hope you’ll enjoy that part too.

The third part I want to talk about is the written assignment. So that’s very different from your normal sort of essay assignment. It’s actually a blog post format. And it’s worth 30% of your mark.

So at the intensive school, we have a scaffolding workshop about that where you get to write a first draft and then you get to peer review each other’s drafts. So in general, people find that very valuable. I don’t require APA formatting for the blog post. You can use any kind of referencing formatting you like.

I find that picking on APA formatting tends to distract from reviewing the content of the post. People find that quite freeing. Some people do find it a bit confusing. You can use APA format if you want of course, you just don’t have to. This can be a challenge for a lot of people, but it’s a useful exercise in trying to write something for the general public while still using your critical thinking faculty.

And the other thing about that is the discussion board posts are really quite a good warmup for that, ’cause they get you thinking about things and writing them in a slightly informal style.

Then the one we all dread, the final exam - that’s worth 40%. That’s also multiple choice and short answer format.

And it’s all online as a MyLearn quiz rather than a supervised exam, which I try and avoid as much as I possibly can. I’ll never do one if I can avoid it cause I find them quite invasive. Don’t tell the powers that be I said that. So I expect you not to cheat. I expect you not to use chat GPT to answer the short answer questions, but you won’t get a good answer if you do.

I’ll make them such that ChatGPT would find them pretty difficult to do. So don’t do that. But if you want to learn something also you’ll think it through for yourself and write your own answers. People tell me it’s challenging, but not impossible. But don’t worry about that, right now, that’s ages away. That’s not till the end of trimester.

Readings etc.

• These are on the Unit Outline (please read it before posting any questions on MyLearn) and also set out in each Topic
• Textbook: Research Methods in Psychology (4th Edition) (open textbook)
• Secondary text: A Student’s Guide to Open Science
• Specific chapter readings are set out for each week
• There will also be other weekly readings from various sources
• I divide the readings into Required and Recommended/Extra
• The Extra ones are not examinable but should give you more depth if required

All right, so readings these are on the unit guide please read that before posting any questions on MyLearn. I know you will, of course. And they’re also set out in each topic, so it’s pretty hard to miss what the required readings are. So each week there’ll be at least one required reading and there are also suggested readings.

You absolutely don’t have to read the suggested readings. In fact sometimes they’re quite a lot more than I expect any one person to be able to do. Those are just to extend your knowledge if you’re interested to give you a bit more rounding around the topic. Sometimes they’re podcasts. So I like to present different formats for things as well, so I hope you’ll enjoy those too.

The textbook is Research Methods in Psychology, fourth edition. It’ll be hard to miss that because the readings are linked actually directly from the MyLearn site as well. If you have the third edition, it’s pretty similar. But you can easily get the fourth edition just by going to the website and clicking on the chapter readings set out for each week.

And there’ll also be other weekly readings from various sources, as I said earlier. A secondary book I’ve added this year is a new one, just published last year. It’s a book called A Student’s Guide to Open Science. And in this book, the author goes over many things that I’m going to talk about to you in this course.

So, while this a bit of extra required reading, it’s not a hard book to ready - I downloaded a Kindle edition, and I read it all the way through. It’s very readable. So I highly recommend this book if you wanna learn a bit more about open science and how to do it well. That’s by Charlotte Pennington, is the author.

All, these links are live. So they should work in the HTML or in the pdf as well.

Intensive school - April 2024

• Held online from 17th - 18th April (mandatory)
• Concurrent with PSYC 421
• There are NO ASSESSMENTS AT INTENSIVE!
• Also no formal lectures - set of interactive workshops and tutorials
• Includes “scaffolding” workshops for assignment.

Okay. So the intensive school, this is in April. It’s held online on Zoom from the 21st to the 22nd. And it’s mandatory. You have to attend. Obviously, it’s most unfortunate if people can’t attend, but because it’s online, it’s not that hard to attend.

We used to have an in-person one. And I miss that a bit, but it was expensive for people to come to Armidale and we generally had a very strong response that people would really prefer to have it on Zoom. So that makes it not particularly difficult for you to attend. It’s held concurrently with Psych 421, so 421 is on the 19th and 20th and 422 is on the 21st and - the 22nd is actually Saturday. So just be aware of that. If you have any childcare issues or anything like that hopefully you can get around those. That’s just how they scheduled it this year, I’m not really quite sure why, but that’s what happened. There will be no assessments at the intensive, of course.

There used to be! They used to sit you down in an exam room and make you do a multiple choice quiz. And of course the minute I took over the unit, I cancelled that, ’cause why would you wanna do that? So there’s no formal lectures either. There’s a set of interactive workshops and tutorials.

And I hope you’ll find them fun and educational and they’re a little bit different from what we do in the online stuff. I want to make the most of being able to interact with you online. So we have breakout groups and that kind of thing. We also have these scaffolding workshops for the assignment where you do a little quick draft of your blog post assignment and then you peer review each other’s drafts.

Most people find that extremely constructive as well, and you tend to get really useful comments - I do read through them and people are great. So that’ll help you write a better assignment, I hope. I’ve also slightly changed the rubric for the assignment on the advice of my markers to make it clearer. So this year that’s just very slightly different. I hope you’ll find that slightly clearer.

Things to note (2024)

• Make sure you have the right edition of the open textbook (available as a paper book on Amazon too)
• NEW: A Student’s Guide to Open Science by Charlotte Pennington - required text (but the library has it)
• “Flipped classroom” model for Intensive School (no lectures)
• All quizzes online, best 5 count
• Half knowledge, half stats
• You don’t need SPSS though! (jamovi is fine - you’ll pick it up quickly, very similar and free)
• There are a whole set of H5P learning modules for jamovi on Moodle if you want them! (in the Resources section)

kay, so things to note for this year. As I said, there’s a new edition of the open textbook - newish - but it’s also available - the reason I chose this one is it’s available as a paper book on Amazon. That link here is to Amazon Australia.

So you can order it as a paper book if you really want. I don’t think it takes that long to arrive because it’s Amazon AU - that’s 20 something bucks or something. So if you really wanna have a paper book to page through, you can totally run with that.

Also new is the Student’s Guide to Open Science that I mentioned before. This is the only required text that costs money, but the library also has a copy and so you can read it online for free if you like.

We have a flipped classroom model for the intensive, so we don’t have lectures. That was something that I got in my feedback from the first year that ran the intensive school with lectures and everyone said look, we can watch lectures online. Why do we come to intensive school just to watch a lecture? And I was like fair enough. So everything we do is interactive and you go off in groups and do stuff. You’ll find out more when you get there. So I hope you’ll enjoy that too.

All the quizzes are online and the best five count. I said this before.

And the other thing I really want you to know is you don’t need SPSS for this course. We do stats, but I’m going to show you how to use jamovi, which is a free statistical software that you can download. It was actually originally designed especially for teaching stats. So it’s very simple. It’s really easy to use. If you can use SPSS, you can totally use it.

There’s a couple of things that are slightly different, but I’ll go through them with you today in the lecture.

And also I’ve made a whole set of H5P learning modules which are on Moodle and you can work through them. They’ve got little questions to test yourself. It’s in the resources section. I spent quite a bit of time last year doing those, and then this year I’ve changed them over to a new set of exercises based on some data about people’s reactions to climate change. You can work through them many times as you like, and I hope you’ll find them helpful as well. But I do feel free to use SPSS if you want.

What I generally find is that people who are more uncertain about stats tend to want to stick with what they know. And, when they eventually change over, they’re like, why did I not use jamovi? It’s so much easier. Another, good reason for using SPSS might be that your supervisor wants you to - maybe you are doing your Honours project, and your supervisor strongly recommends that you use SPSS - there may be analyses that you’re doing that aren’t available in jamovi, for instance. If you’re doing anything with missing data replacement, that kind of complex thing, SPSS does do a lot more of that stuff. So there are plenty of valid reasons for using SPSS and I don’t want to totally diss it, but even if you are doing that, you might want to try jamovi anyway, because you might find that you get more of an understanding of the underlying principles behind the stats. By using a program that’s somewhat simpler or slightly different, it becomes less about using SPSS and more about actually doing statistics. Bearing that in mind, you can even use other programs like PSPP, or you can use JASP or if you’re really nerdy, you can use R.

That’s what I use, in fact. But most of what I’m going to be telling you about today is jamovi. Yeah, stick with me.

My teaching philosophy

Psychology as a science

Science makes systematic observations

Does Psychology?

• Similar to knowing things from experience, but more systematic
• Including:
  • Precise definitions
  • Reliable and valid measuring tools
  • Generally accepted methodologies
  • A system of logic for drawing conclusions

Science also bases its findings on observations, but they are made much more systematically than our everyday observations.

The scientist’s systematic observations include using:

• Precise definitions of the phenomena being measured
• Reliable and valid measuring tools that yield useful and interpretable data
• Generally accepted research methodologies
• A system of logic for drawing conclusions and fitting those conclusions into general theories.

Science produces public knowledge

Does Psychology?

• The objective scientist was believed to be almost machine‐like in the search for truth.
• In science this usually takes the form of defining terms and research procedures precisely enough so any other person can systematically repeat the study, presumably achieving the same observable outcome.
• Objectivity in psychological science has been a problem historically.

This is the idealistic view that your textbook sets out. Objectivity - You’re not kidding! We’re going to go into this in a LOT more detail in this course. This second point, in particular, is a doozy. Most published papers, for a start, are currently NOT publicly available, and do NOT contain precise enough descriptions that any other person can repeat the study. This has resulted in what we now call the Replication Crisis in Psychology, and you’re going to hear all about that in my next lecture.

Science produces tentative conclusions

Does Psychology?

• Related to data‐driven attitude: recognition that conclusions drawn from data are always tentative & subject to revision based on future research.

• Tentative nature of scientific research often difficult for general public

• People seem to believe outcomes of well‐executed research will be the authoritative & the “final answer”

Also a doozy of naivety here. Slide is too wordy - revise

BUT: James Heathers tweet

And a related thread here…

Science asks answerable questions

Does psychology?

Empirical questions can be answered through the systematic observations and techniques that characterise scientific methodology.

Science has theories that can be falsified

Does psychology?

• Empirical question -> hypothesis (prediction about outcome)
• A good theory must be precise to be disproven, at least in principle.
• This concept is often referred to as falsification.

• When designing research studies, an early step in the process is to reshape the empirical question into a hypothesis, which is a prediction about the study’s outcome.
• A critically important attribute of a good theory is that it must be precise enough so it can be disproven, at least in principle.
• This concept is often referred to as falsification.

BUT: HARKing

• “Hypothesising After Results are Known”
• Why is this a bad thing?
• The Texas Sharpshooter Fallacy (see also this post)
• Increases risk of Type I error (false positives)
• Original article (Kerr, 1998!)

Remember to link article PDF to current moodle site.

One of the problems when we try to draw understanding from published papers is the practice of HARKing (Hypothesising After Results are Known). This is what happens when you collect a bunch of data, then look for patterns within the data and then write your introduction as if you had predicted those results all along! (Of course there’s nothing wrong with exploratory analysis of your data, as long as you’re very clear that that’s what you are doing.) This is like the legendary Texas Sharpshooter who sprays bullets randomly at a fence and then draws targets around the bullet holes. Sadly, you will sometimes see this practice encouraged even today (for instance, in the Bem article that is often given as an example of how to write a paper). We’ll be talking more about Daryl Bem later in the course.

Also but:

Does psychology have good theories?

• Psychology has long been criticised for being weak on theory
• See, for instance, Muthukrishna, M, and J Henrich. “A problem in theory.” Nature Human Behavior (2019).
• A really good Twitter thread on this topic - see below
• Eiko Fried on Twitter

So thinking back to our previous slide, the hypothesis or hypotheses you have formed should have arisen from good theories. But does psychology really have strong theories? Sometimes it has been criticised for finding effects rather than forming theories. I found a really good thread about this on Twitter this morning - here’s a link if you want to follow it. But this is a bit of a distraction from my main point here, which is about the falsifiability of your theory or, more narrowly, your hypothesis.

Recognising pseudoscience

We can, right?

• Associates with true science
• Reliance on anecdotal evidence
• Sidesteps the falsification requirement
• Reduces complex phenomena to simplistic concepts

But getting back some of the content from your textbook. How can we recognise pseudoscience in everyday life? A lot of you have been already posting some great examples on the Discussion board for this week and I’m greatly enjoying them. [NOTE: UPDATE WITH THIS YEAR’S POSTS] Reiki, energy healing, the dangers of 5G and COVID-19 vaccinations, are just a few we’ve seen, as well as some examples of good science reporting, which is really nice to see. Another more impressive example that came up last year is the rise and fall of Theranos, a very impressive and sciencey-sounding tech startup that turned out to be based on an idea that was pure pseudoscience.

Recognising pseudoscience

We can, right?

• People in the late nineteenth century could send away to the New York firm of Fowler and Wells for a ‘Symbolic Head and Phrenological Map’ for 10 cents.
• For another $1.25, the head and map would be accompanied by a copy of How to Read Character: A New Illustrated Handbook of Phrenology and Physiognomy.
• Of course, we don’t believe that sort of thing any more - do we?

Science vs. pseudoscience

Here’s a replica of that bust you could buy for 10 cents. Lots of psychology academics like to keep one of these in their office, perhaps as a reminder of the foolishness of past times. But are we really less gullible now, even as psychologists? Here’s a nice little article from Time magazine covering a fairly recent paper about the use of machine learning to diagnose autism. I did a reverse image search on the image they used to illustrate this article - it isn’t an image from the article, nor is it even an image from an fMRI study at all (which I happen to know because I’ve done some fMRI). In fact it’s a stock photo from the science photo library which seems to be used to illustrate almost any article about something with brains in it, which you’ll discover if you do the reverse image search yourself.

Papers & criticism

• Article in Time Magazine
• Original article in PLoS One
• Another paper about machine learning and autism
• A thoughtful critical piece by Jon Brock
• What do you think?

But is the article itself OK? It was published in PLoS One, which is a reasonable open access journal (disclosure: I’m an editor there), although it has published a few sketchy things from time to time. It was published in 2014, and the first thing I’d like to say is that machine learning from fMRI data has moved on quite a lot since then - it’s a field that moves pretty quickly. The small sample size would lead me to be cautious of overfitting (the accuracy of 97% seems too high) in this instance, and in the links here you can see some thoughtful criticism of this type of study. See what you think! But the headline is definitely very sensationalist.

A more recent example

• Hydroxychloroquine anyone?
• Slate article here
• Pubpeer commentary here
• And here (on published version)
• Later RCTs showed it was ineffective
• But damage had been done

Has published 3000 scientific articles. Cultivates politicians & created his own research institute. Announced the results of this study in a YouTube video. Trial was not double-blind, participants dropped out, impossible to interpret result. Note: Update with more recent example?

Another example: Ivermectin

• A good example of how misinformation spreads
• Desperation (pre-vaccine) led people to try many things
• A good article here about the Ivermectin story
• Many studies including clinical trials, but some larger studies later shown to be fraudulent (see Health Nerd on Twitter for more details)
• Meta-analysis looked good at first
• But garbage in, garbage out - later retracted.

Practical section (stats demos and resources)

FREE alternatives to SPSS

• jamovi (good for parametric and non-parametric stats - lots of add-ons including moderation/mediation - based on R)
• JASP (good for Bayesian statistics)
• Posit (formerly R Studio) (very powerful & flexible, fairly steep learning curve, great plots)
• PSPP - most SPSS-like but clunky

• Update R Studio link to latest
• include logos?

Learning resources

All free!

• Fantastic FREE book by Danielle Navarro about learning statistics with jamovi (more in here than just jamovi - some great info about stats and research methods)
• Online FREE course on jamovi with video tutorials (by datalab)
• For SPSS - YouTube is great!
• Statistics of DOOM YouTube Channel (thanks Bernie!)

Stats demo example

(from my own data)

Force plate for measuring postural sway

Sway occurs in the mediolateral (ML) and antero-posterior (AP) directions

Stats demo example

(from my own data)

• Force plate measures Centre of Pressure (CoP) as it moves about the centre of balance
• People generally sway more with eyes closed than open

Why not include the animated plots here???

Research on motor control in Parkinson’s disease

• Parkinson’s disease affects brain structures also implicated in postural control (cerebellum, brainstem, basal ganglia)
• Postural control is known to be affected (increased sway) in PD
• Our group is interested in tracking this across time to see if if can be a useful measure of disease progression
• We also measure cognitive functioning and use some more traditional clinical tests

Paper - now published in BMJ Neurology Open

• Apthorp, D., Smith, A., Ilschner, S., Vlieger, R., Das, C., Lueck, C. J., & Looi, J. C. L. (2020). Postural sway correlates with cognition and quality of life in Parkinson’s disease. BMJ Neurology Open, 2(2), e000086. https://doi.org/10.1136/bmjno-2020-000086
• Data and code all available here
• We’re going to use a slightly reduced version of the dataset to do some demos in jamovi - I’ll put that dataset up on MyLearn

Demo: Intro to jamovi

Demo: Eyes open vs. closed

Demo: Comparing two groups

Demo: Filtering in jamovi

Demo: Correlations and scatter plots

Questions? Ask on MyLearn!