Sampling

Question 1

Universe

Answer 1

Broad population to which you want to generalize your findings (e.g. everyone with anorexia)

Question 2

Population

Answer 2

The defined group within the universe from which the sample will be selected (e.g. anorexia patients across 10 hospitals)`

Question 3

Sample

Answer 3

The participants who actually take part in the study (e.g. 50 participants from those hospitals)

Question 4

Statistic vs. population parameter

Answer 4

Statistic: quantitative measurement from a sample (what we actually get in research)

Population parameter: estimate of the value in the population (what we want to know but almost never know with certainty)

Question 5

Confidence interval

Answer 5

A range of values for which are plausible for the population

A generalized value for the population

Question 6

Threats to external validity

Answer 6

Sample characteristics, ecological validity, reactivity, multiple treatment interference, novelty effect, disruption effect, test sensitization

Question 7

Sample characteristics (EV threat)

Answer 7

Differences between study sample and other samples you wish to generalize the findings to (age, gender, etc)

solution: get a representative sample

Question 8

Ecological validity (EV threat)

Answer 8

The methods, materials, and setting of the study must approximate the real-life situation under investigation

solution: try to make research as close to real-life as possible

Question 9

Reactivity of experimental arrangements (EV threat)

Answer 9

Awareness of being in a study may affect behaviour or elicit certain reaction; participants have theories about research/may act the way they think you want them to act

solution: don’t let participants know hypotheses beforehand

Question 10

Multiple-treatment interference (EV threat)

Answer 10

If many treatments are applied, it is difficult to determine how well each of the treatments would work individually

solution: treatment 1, treatment 2, control, both

Question 11

Novelty effect (EV threat)

Answer 11

Treatment is only working because its ‘new’; people are enthusiastic and expect it to work

solution: wait a while after the program has started before starting evaluations

Question 12

Disruption effect (EV threat)

Answer 12

Treatment appears to fail in the short term because it has disrupted daily life/routine

solution:wait a while after the program has started before starting evaluations

Question 13

Test sensitization

Answer 13

Does pre-testing or the test itself alter subject experience/response (ex. what effect does rating your mood prior to a mood induction technique have your mood)

solution: avoid within-subject design, or multiple condition varying the order or stimuli presented

Question 14

Sample selection bias

Answer 14

Convenience sampling (admitting anyone you can find into the study from a population that’s easy to recruit)

WEIRD research
80% of participants were psych undergrads
96% cone from countries that make up 12% of the population

Question 15

Attrition (EV threat)

Answer 15

Longitudinal design issue

Number of dropouts across groups could effect results

solution: reminders, backloading

Question 16

Statistical power

Answer 16

how many participants are needed to have a successful experiement

Question 17

Statistical power relies on…

Answer 17

1. Effect size
2. Alpha (type I error)
3. Beta (type II error)

Question 18

Effect size

Answer 18

How large is the effect?/What is the magnitude of the effect?

Question 19

Measuring effect size

Answer 19

Pearson’s r (.1 = small; .3 = medium; .5 = large)

Cohen’s d (.3 = small; .5 = medium; .8 = large)

Question 20

Alpha

Answer 20

Type I error; false positive = 0.05

probability of finding an effect size when none exists

Question 21

Beta

Answer 21

Probability of missing an effect that IS present

Type II error

Usually .80 (20% chance of type II error)