Which Method
Would You Use?
Ten research questions. Five tools. Turn and talk, commit to a method, and be ready to defend it.
Experiment
Correlational study
Survey
Naturalistic observation
Case study
AP Psychology · Unit 1 · Research Methods Review
How this works
1
Read the question.
What is actually being asked — a cause, a relationship, a description, or a depth?
2
Turn and talk.
Thirty seconds with a partner. Name a method and say why.
3
Commit.
Everyone picks one. No abstaining.
4
Defend it.
Say what your method can establish — and what it cannot.
The question behind every question
Did the researcher change something?
Did chance decide who got what?
If no to either, it is not an experiment — and no causal claim is available.
1
REACTION TIME
Does caffeine improve reaction time?
You can give it to some people and not others, and you can decide who gets which. What does that let you claim?
STRAIGHTFORWARD
2
AFTER A STROKE
A patient can no longer recognise faces — but recognises voices perfectly.
One person. An injury nobody could ever create on purpose. What method is even available?
STRAIGHTFORWARD
3
SLEEP
What percentage of students at this school get eight or more hours of sleep?
You need a number that describes thousands of people. Nobody is being manipulated and nothing is being caused.
STRAIGHTFORWARD
4
THE PLAYGROUND
How do five-year-olds resolve conflicts over toys when no adult steps in?
The moment you intervene, the thing you wanted to study stops happening.
STRAIGHTFORWARD
5
SCREENS AND GRADES
Is there a relationship between hours of screen time and GPA?
Both already vary. You are measuring, not assigning. What can you say when you find they move together?
STRAIGHTFORWARD
6
VIDEO GAMES
Do violent video games cause aggression in teenagers?
The word “cause” demands an experiment. The ethics of the study you would need make that hard. So what do you actually do?
MORE THAN ONE DEFENSIBLE ANSWER
7
A NEW THERAPY
Does a new therapy reduce symptoms of depression?
You could randomly assign it. Should you? What happens to the people in the control group?
MORE THAN ONE DEFENSIBLE ANSWER
8
LANGUAGE
How does language develop between ages one and five?
You could follow the same children for four years. Or you could test one-year-olds and five-year-olds next Tuesday.
MORE THAN ONE DEFENSIBLE ANSWER
9
JONESTOWN
Why did more than 900 people die in a single day at Jonestown?
No experiment could ever produce this. But a case study cannot generalise. So what is the method actually doing here?
MORE THAN ONE DEFENSIBLE ANSWER
10
CHILDHOOD POVERTY
Does poverty in childhood affect health in adulthood?
Write the experiment. Actually write it out — what would you have to do to the participants?
MORE THAN ONE DEFENSIBLE ANSWER
One method answers
“what caused this?”
Every other method can find a relationship, describe a population, or go deep on a single case. None of them can tell you what caused it — and for a great many real questions, that is the only evidence anyone will ever have.
Next
Descriptive statistics — what to do with the numbers once you have collected them.