Basic Stats

Question 1

What central measure of tendency should we use for nominal data?

Answer 1

Mode

Question 2

What measure of central tendency should we use for ordinal data?

Answer 2

Median

Question 3

What measure of central tendency should we use for interval data?

Answer 3

Either mean or median

Question 4

What measure of central tendency should we use for ratio data?

Answer 4

Either mean or median.

Question 5

What is the kurtosis of a tall skinny distribution?

Answer 5

Leptokurtic! Cause it leaps :D

Question 6

What is the name of a short fat distribution?

Answer 6

Platokurtic… like flat.

Question 7

What is CLT?

Answer 7

Central Limit Theorem. It suggests that the bigger the sample, the more the mean distribution will be normal.

Question 8

What measure of variability should we use for nominal data?

Answer 8

None

Question 9

What measure of variability should we use for ordinal level data?

Answer 9

Range

Question 10

What measure of variability should we use for interval level data?

Answer 10

Variance or SD

Question 11

What measure of variability should we use for ratio data?

Answer 11

Variance or SD.

Question 12

A standard deviation is what to a variance?

Answer 12

Variance squared

Question 13

T-scores:
A) used for: ______
B) how does it work: ______

Answer 13

A) when SD of the population is unknown or when sample is 30 or smaller.
B) M = 50, SD = 10

Question 14

What is a z-score?

Answer 14

It is transforming raw scores into units of standard deviations. It makes it easier to compares scores across different scales.

Question 15

What is a Type I error?

Answer 15

A false positive. It is rejecting the Ho (the hypothesis that there is no effect/ no relationship between the variables) when it is in fact true.

Question 16

What is a Type II error?

Answer 16

NOT rejecting the Ho when it is not true, i.e. assuming there is no effect when there is.

Question 17

The variable we manipulate is…

Answer 17

INDEPENDENT

Question 18

The variable that we do not manipulate and, in other words, the outcome variable is…

Answer 18

DEPENDENT

Question 19

Due to the nature of correlational studies, 2 or more … variables are studied. Why?

Answer 19

DEPENDENT. Correlation is not causation, we are not manipulating variables - that is an experiment. We are simply looking at the relationship between 2 or more variables.

Question 20

What is a latent variable? What is the other kind?

Answer 20

A latent variable measures a trait that cannot be seen, touched, etc., such as perfectionism. The other is observed or construct variables.

Question 21

Is temperature ratio or interval? Why can’t it be ordinal?

Answer 21

It is interval as it is not possible to not have a temperature. It can’t be ordinal, because when transformed to Fahrenheit, the distance between 20º and 40º is no longer of 20º (68º to 104º respectively).

Question 22

Publication bias - explain. Give an example of a counter action.

Answer 22

Publishing only studies with statistically significant results. A way to get around this is by pre-registering studies and publishing ‘regardless’ of results.

Question 23

What kinds of replication can be done for a study?

Answer 23

Direct (replicating exactly per the original) or Systematic (changing one of the conditions, i.e. a different population)

Question 24

What is inductive reasoning? How is it different from deductive reasoning?

Answer 24

Inductive reasoning by reasoning from specific instances to a generalisation. Example: I’ve watched a few romantic comedies and they’re shit, so all of them must be shit.
Deductive reasoning is when a conclusion is deduced from a general rule. Example: All romantic comedies are shit, so this one must be as well.

Question 25

Give an example of ordinal data.

Answer 25

The scale of self-esteem is ordinal, as ‘strongly agree’ is stronger than ‘agree’.

Question 26

Give an example of ratio data.

Answer 26

Age.

Question 27

What is the cohort effect? And how does it affect longitudinal studies?

Answer 27

People change from generation to generation. This will be seen in the scores. We could do a study on the effects of ChatGPT on children’s ability to write but it would probably be made obsolete by some other technology by the time it gets approved by ethics.

Question 29

This isn’t a question. Just think which is stronger between +.11 and -.89. And remember that the sign is just a symbol DIRECTION, not STRENGTH.

Answer 29

The one that is stronger is .89, because it is closer to 1. Closer to 0 is WEAKER.

Question 30

One-tailed hypothesis versus two-tailed:

Answer 30

1-tail is for when you have an expectation of the direction of the relationship. 2-tail is when you don’t.

Question 31

What is the Fundamental Attribution Error? Now called the Correspondence Bias:

Answer 31

It is attributing someone’s actions to their traits or characters while disregarding environmental factors that might have an effect. (Eugenecists do this).

Question 32

What kind of correlation do you use for each out of the 4 types of data?

Answer 32

Nominal - none Ordinal - Spearman’s Interval or ratio - Pearson’s r