Methods in Psychological Science

Question 1

Theory definition

Answer 1

An attempt to explain natural phenomena
- generates testable predictions

Question 2

Hypothesis definition

Answer 2

A specific (and falsifiable) prediction made by a theory

Question 3

Rule of parsimony

Answer 3

Simplest theory that explains all the evidence is the best one

Question 4

Why is psychology harder to study than physics?

Answer 4

• Humans are complex (in physics you can observe something physically, but in psychology, it is very hard to observe what goes on inside us in our brain, leading to different behaviours)
• Humans are variable (no two people ever behave in the same way; the same person behaves differently now than they did many years ago)
• Humans respond to being observed (demand characteristics, the Hawthorne effect)

Question 5

What is meant by “operationalizing the hypothesis”?

Answer 5

Define what you’re interested in (should have construct validity) + create a way (an instrument) to measure and test in concrete, objective terms (should have reliability and power)

Question 6

What is a psychometric and what are some key features of a good psychometric?

Answer 6

A psychometric is a test or technique used to measure psychological attributes (intelligence, aptitude, etc.). Features of a good psychometric:
- power
- reliability
- validity

Question 7

Define power

Answer 7

Sensitivity to detect small changes (e.g., 9.58s vs 9.59s)

Question 8

Define reliability

Answer 8

Tendency to produce the same result consistently (result stays stable over time)

Question 9

Define validity

Answer 9

Extent to which a measurement and a property are conceptually related. Asks: does your tool measure what you think it measures? For e.g., to objectively measure the property of distress in a baby, we can measure the length of time for which it cries- has some construct validity and what is being measured and the property we’re interested in are conceptually related.

Question 10

Steps of the scientific method

Answer 10

Identifying a problem
Generating theories and hypotheses
Designing the study
Data collection
Using the data to test your hypotheses
Reporting your findings (diff course)
Replication and open science practices

Question 11

Common research designs in psychology

Answer 11

• single case study (in neuropsychology, e.g., Broca’s study of Mr. Leborgne)
• correlational study (often using questionnaires)
• naturalistic observation (nothing is manipulated, simple observation)
• experiments

Question 12

Elements of an experiment

Answer 12

1. Manipulation:
   Subjects
   Independent variable- manipulated
   Dependent variable- measured
   Control group
2. Sampling
   Random assignment (avoid self selection)
   Controlling for important subject variables (confounds)
   Be aware of convenience samples (may not be representative of the general population): students are WEIRD (westernized, educated, industrialized, rich, democratic)

Question 13

What is the Hawthorne effect?

Answer 13

Participant performance changes when they feel watched

Question 14

What are demand characteristics? How can researchers avoid these?

Answer 14

Participants behave as they think they should. To avoid this researchers use:
- covert measures
- deception (debrief after study)

Question 15

Observer (experimenter) bias

Answer 15

We interpret data to match our expectations. To avoid this:
- use double-blind designs (both the experimenter and participant are blind to what is being observed)
- standardize data analysis

Question 16

Why use statistics to analyze data?

Answer 16

• to summarize + organize the sampled data (descriptive statistics)
• to interpret whether differences in sampled data are likely to be meaningful for a larger population (inferential statistics)

Question 17

Descriptive statistics: measures of central tendency

Answer 17

Mean: sum/n
Median: middle value (50% of values above and 50% below this value)
Mode: most frequent observation(s)
Outliers: odd/uncharacteristic observation(s)

Question 18

Descriptive statistics: measures of variability/dispersion

Answer 18

Range: difference between min and max values
Standard deviation: measure of dispersion from the mean (how far is each datapoint from the average?)

Question 19

Histogram

Answer 19

Plots the frequency distribution of a single variable
Groups data into ranges and displays the frequency of datapoints within each range

Question 20

Measures of central tendency when the distribution is positively skewed

Answer 20

Positively skewed = most of the distribution is to the left of the mean
So mode and median are lower than the mean

Question 21

Measures of central tendency when the distribution is negatively skewed

Answer 21

Negatively skewed = most of the distribution is to the right of the mean
So median and mode are greater than the mean

Question 22

Measures of central tendency for a normal distribution

Answer 22

All values are equally distributed around the mean value
So the mean, median and mode are all the same

Question 23

Key factors in inferential statistics

Answer 23

of people in the sample

• more people reduce the likelihood that our sample is uncharacteristic of the population

Variability between each group
- bigger differences between groups may reflect a stronger effect of your manipulation

Variability within each group
- high variability within a group reduces potential differences between groups. As a result, the data may not generalize to the larger population

Question 24

Inferential statistics produce a p value. What does it mean?

Answer 24

The p value gives the likelihood that the effect of the independent variable (results) are due to chance. We are typically willing to accept 5% or less likelihood that the results are a product of chance (p<0.05)

Question 25

Type 1 error

Answer 25

Claim there is a true effect but it’s just random noise. E.g., telling an old man they are pregnant.

Question 26

Type 2 error

Answer 26

When you miss a true effect that’s really there (maybe you didn’t test enough people). E.g., telling a pregnant woman she is not pregnant.

Question 27

The Nuremberg Code (1947)

Answer 27

Specific to medical research - informed consent - human research based on animal work - benefits > risks - minimize discomfort and avoid injury

Question 28

Issues with the Tuskegee Syphilis Study (1932-72)

Answer 28

Aimed to study the progression of syphilis in 600 black men, who were not informed of the treatment when the cure to syphilis was found. - lack of informed consent - deception - withholding care and information - exploitation of a vulnerable group

Question 29

The Belmont Principles (1979)

Answer 29

Written by a US bioethics commission following Tuskegee 1. Beneficence (welfare): “do no harm”, benefits must outweigh risks 2. Respect for people: respect for autonomy, need for informed consent 3. Justice: equality in participation in the research process (diverse sample of participants so that the results of research can benefit everyone), avoiding exploitation of vulnerable groups

Question 30

Milgram’s Obedience Study (1963)

Answer 30

- motivated by Nazi Germany’s behaviour to understand why people behaved cruelly the way they did under an authority even if it went against their morals - advertised as a learning experiment - the participants thought that learners were getting higher level shocks when they made errors - the study tested how far people went to obey an authority (the experimenter) Issues: - deception - unanticipated psychological harm

Question 31

Research Ethics Boards (REBs)

Answer 31

- study justification, protocol, full study materials (e.g., consent form, ads, protocols) are reviewed by an expert panel - any changes must be approved (an amendment) and renewed annually - in Canada, all members of research team must have TCPS2 training (built on the Belmont principle)

Question 32

Rules for animal research in Canada

Answer 32

- discomfort and stress are minimized - there are benefits to humans and animals - alternative procedures are unavailable (computer models, cell cultures, etc.) - must be conducted by trained scientists

Question 33

Animal Research statistics

Answer 33

- 91% of NIH animal research (USA) uses mice and rats - increasing use of zebrafish and Drosophila (fruit flies) - the majority of studies (nearly 70%) involve no or minor/short-term discomfort

Question 34

Three R’s of Animal Research

Answer 34

Reduction: reduce the number of animals used in experiments as much as possible Refinement: improve the way animals are cared for, minimize pain and suffering Replacement: replace experiments on animals with alternative techniques

Question 35

Confirmation bias

Answer 35

Humans are biased towards information that confirms their existing beliefs. They ignore evidence that might disconfirm their beliefs. E.g., echo chambers on social media. Sir Francis Bacon: “The human understanding, once it has adopted opinions…draws everything else to support and agree with them…”

Question 36

Problems associated with the process of research

Answer 36

- inconsistent methods/lack of transparency - HARKing: Hypothesizing After Results are Known - publication bias: harder to publish null results