Hello,
As we all know, psychology students have a prejudice towards statistics...
What can I do to take their attention and to break this prejudice?
Any interesting class activities???
Thank you in advance...
Hello,
There is a website ("seeing theory") that I find especially attractive because it shows visually a lot of basic stats concepts that are sometimes hard to grasp by looking at the theory. It is easy to use and interactive. Hope you and your students will enjoy it as well !
http://students.brown.edu/seeing-theory/index.html
Cheers !
I think you may need some practical examples to gain their interest.
This book may interest you and your students:
Andy Field-Discovering Statistics Using IBM SPSS Statistics-SAGE Publications
I let them use data that they're interested in. For example, for a ttest, let them compare football teams across leagues, whther dramas or action movies make more money, or collect their own data on whether men or women drink more diet drinks. The key is to have them see that stats is relevant to their questions. Thus, for outside exercises, they pick their own data.
For in-class exercises I might give them data relating GRE score to grad school admission, or how grade on the first test predicts grade on the second. Again, it's relevant to them. They WANT to know the answer!
Good luck!
Inspiring them with unharnessed power to discover objective laws of natural world and convincing them, that statistics is all that we have to achieve this goal may not go so easily, so here's what I recommend :)
- Use the least amount of numbers and mathematical formulas,
- Try to use many examples for abstract terms, eg. what is correlation ?- Its answer to question how often does one thing co-occur with other, when "O" means, they go completely separate ways and 1, they always go together. Then you can throw some specific example, relevant to their experience,
- Try to make sure, whether they have firm understanding of basics- especially, do they know what distribution, central tendency and dispersion are ? They may know the words, but if they don't personalize the meaning, they may have very serious difficulties moving on,
- Try to remember, that they are mostly humanists, so they like and understand words and stories. At firs sight, they may feel unpleasant with symbols and calculations- show them, that they tell the stories as well and they have meaning, relevant to real world.
- Try always to show them meaning of this knowledge, how it can help understand facts and laws of human psyche ? How it can help to understand scientific literature, relevant to their interests ? It's especially important to convince them, that statistics it's not just for scientists. It's something that can make a difference between scientific literate and illiterate professional. So actually, between professional and fraud or amateur.
Jamal Munshi has a paper on the The Spuriousness of Correlations between Cumulative Values
Article The Spuriousness of Correlations between Cumulative Values
What Munshi does is generate two sets of a hundred or so data points using pseudo random number generators. Then he calculates the correlation coefficient which he displays are around zero. At the same time he cumulates the numbers generates and calculates the the correlation coefficient at around 0,999.
I think that students with laptops and spreadsheet programs could easily replicate Munshi's paper.
I don't know what lesson Munshi teaches with this, but I would introduce Granger and Engle,. random walks, the analogy of the drunkard's walk with a dog on a leash.
This would lead to a general discussion of stationarity, mean reversion and deterministic trends, cointegration and Granger causality. Plus warnings about spurious correlation.
Actually, as I recall, when Grander gave his acceptance speech to the Nobel committee, he used the analogy of two loosely joined stings of pearls thrown on a table to form random patterns that were loosely correlated. I prefer the analogy of drunkard with a dog on a very elastic leash. But I don't teach statistics anymore.
I second Michael Young’s answer and raise it one. Got them to collect the data. You don’t need big projects, and you don’t need to turn it in to an APA style paper, just get them to come up up with an interesting question, and go collect the data. it works great if it comes out of a class debate... Who hates math more, men or women? Would stats be better if it wasn’t a morning class? Did their psych professors like stats when hey undergraduates more than they do? Send out to get a few data points between classes, and the next class period they can use the data to answer their own question. My guess is that more of them will be engaged, and most of them will get a better sense of what it all means.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics