Week 2 · Chapter

Chapter 2 — Research Methods & Ethics

Introduction to Psychology Free sample edition

The module's primary reading · ~35–45 minutes · Objective 2 (evaluate psychological research methods and ethics, distinguishing correlation from causation, random sampling from random assignment, and identifying sources of bias) · Read it first this week; the lecture, the tutorial, the quiz, and the Analysis Workshop all build on it.

In plain language

Psychology makes claims about people, so it needs a fair way to check them. This week you learn the main ways psychologists collect evidence and how much each way is allowed to claim. A correlation (two measures that tend to rise or fall together) can show that two things are linked, but not that one causes the other. An experiment (a study where the researcher changes one thing on purpose and watches what happens) can show a cause. You will also learn the rules that protect the people who take part in a study.

Why you already care about this

Ten students in a campus study group compared notes one morning. Each wrote down how many hours they had slept and their score on that day's 20-point quiz. The students who slept about 8 or 9 hours mostly scored 14 to 18. The student who slept 5 hours scored 11. A headline writer would love this. "Sleep raises quiz scores!" But is that what the numbers show?

You meet claims like this every day. A wellness app says its users feel calmer. A news story says people who walk to work are happier. A friend says a supplement helped them focus. Each claim sounds like it came from evidence. The question this chapter teaches you to ask is: what kind of evidence, and what is it allowed to say?

Psychologists answer that question with a small toolkit, and you can use it too. They define what they measure. They choose a design that fits the question. They check whether the result holds up. And they protect the people who make the research possible.

By the end of this chapter you will be able to:

write an operational definition for a fuzzy idea like memory or stress;
name the design behind a study and say what that design can conclude;
read a correlation coefficient for direction and strength, and name a third variable that could explain it;
label the independent variable, the dependent variable, and the control group in an experiment;
tell random sampling from random assignment, and reliability from validity;
name the ethical protections every study with people must provide.

Carry this sentence with you: a correlation shows a link — not a cause.

Where we are

Already covered: Week 1, The Science of Psychology, defined psychology as the science of behavior and mental processes and showed why common sense and hindsight are not enough.
This week: how psychologists gather evidence, what each research design can claim, and the ethics rules that protect participants.
How they connect: Week 1 said a science tests its ideas against evidence; this week shows exactly how that testing works.
Where we're going next: Week 3, Biological Bases of Behavior, looks inside the brain and the nervous system, and you will read every brain finding there with this week's questions in hand.

Week 1 is the one this chapter leans on. Recall two words from it. A theory is a broad explanation that is already backed by evidence. A hypothesis is a specific prediction that data could prove wrong. Recall also hindsight bias, the feeling that you "knew it all along" once you hear a result. Research methods are the cure for hindsight bias. They make you state the prediction before the data come in.

1. Run the research loop: turn a fuzzy idea into a measurement

Imagine your roommate tells you, "I remember things better in the morning." That is a hunch. It might be right. But "remember better" is fuzzy. Better by how much? Measured how? Until someone answers those questions, nobody can check the claim.

Psychological research runs as a loop. A theory produces a hypothesis. The hypothesis needs a measurement. Data come in. Then the result has to hold up when someone repeats the study. If it does, the theory is kept. If it does not, the theory is revised or dropped.

The step that turns a fuzzy idea into something you can count is the operational definition.

An operational definition is the exact, observable procedure a researcher uses to measure a construct, such as measuring memory as percent savings on relearning; a construct without one is not yet measurable.
Replication is repeating a study, with the same or new participants, to see whether its result appears again; a finding that does not replicate is not kept as settled.

A construct is any idea you cannot see directly, such as memory, stress, or kindness. An operational definition makes it visible. "Reaction time" becomes milliseconds to press a key after a light flashes. "Helping" becomes reporting an emergency within a set time window.

Worked example — Ebbinghaus measures memory. Hermann Ebbinghaus, whose memory experiments were published in 1885, wanted to study forgetting. "How much do I remember?" was too fuzzy. So he memorized lists of nonsense syllables, waited, and then relearned the same lists. He measured memory as percent savings: how much less time or effort relearning material takes than learning it the first time, written as a percentage of the original learning.

The definition. Memory equals percent savings on relearning. That is a procedure anyone can repeat.
The data. Savings fell from 58.2% at 19 minutes to 21.1% at 31 days.
The size of the drop. 58.2 − 21.1 = 37.1 percentage points over the month.
Where the drop happened. From 19 minutes to 63 minutes, savings fell from 58.2% to 44.2%. That is 58.2 − 44.2 = 14.0 points in under an hour. From 6 days to 31 days, savings fell from 25.4% to 21.1%. That is only 25.4 − 21.1 = 4.3 points.

In words: most forgetting happened fast, and then it slowed down a lot. Because Ebbinghaus wrote down his exact procedure, other researchers could repeat it. That is what makes the finding science and not opinion. You will meet this curve again in Week 7, when the course turns to memory itself.

Misconception check. Students often think a construct like "memory" or "intelligence" can be measured directly, the way you measure height with a tape. It is tempting because we use these words every day as if they were solid things. But nobody can see a memory. Every study measures a stand-in: a score, a time, a count. The cure: when you read a finding, ask "how did they measure that?" If the procedure is vague, the finding is too.

One honesty note. A single study is never the last word. A result can come from luck, from a quirk of one sample, or from a mistake. That is why replication sits at the end of the loop.

2. Describe behavior: case studies, observation, and surveys

Picture a researcher you might pass, sitting near a busy four-way stop with a clipboard. She is not stopping anyone or asking any questions. She notes how many drivers come to a full stop and how many roll through. At the end of the week, she has a careful record of what drivers do. She does not know why they do it.

You are looking at the job of descriptive research: research that records what people do, say, or report as it happens, through case studies, naturalistic observation, or surveys, without testing what causes it. There are three main kinds.

A case study is a descriptive method that examines one person or a small group in great depth; it can suggest ideas to test but cannot show cause or be assumed to fit everyone.
Naturalistic observation is a descriptive method in which a researcher watches and records behavior in its everyday setting without trying to change it.
A survey is a descriptive method that asks many people to report their own attitudes, beliefs, or behavior, usually through a questionnaire or an interview.

Each method gives you something useful. A case study can follow one person's rare experience in rich detail. Naturalistic observation sees real behavior, not lab behavior. A survey can reach thousands of people at low cost. Each also has a limitation. A case study may not fit anyone else. People who know they are being watched may act differently. And people answering a survey may misremember, or give the answer that makes them look good.

Misconception check. Students often think a big survey proves what causes something. Suppose a survey of a large office finds that workers who take lunch away from their desks report less stress. It is tempting to say "taking lunch away lowers stress." But the survey only recorded two answers. Maybe less-stressed workers feel free to leave their desks. The cure: a descriptive study describes. To claim a cause, you need a different design.

3. Read a correlation: direction, strength, and the "associated with" sentence

Go back to the ten students you met in the opening. Nobody changed their sleep. The researcher measured two things that were already happening: hours of sleep and the next day's quiz score. That makes it a correlational study.

A correlational study is a design that measures two or more variables as they naturally occur and reports how strongly they move together; nothing is manipulated, so it has no independent or dependent variable.
The correlation coefficient (r) is a number from −1 to +1 that describes a linear association between two measured variables: its sign gives the direction and its size gives the strength, and a value near 0 means little linear association.

Here are the data. They are illustrative: the course built them to teach with, so they are not from a real study.

Hours of sleep and next-day quiz score (out of 20) for ten students, an illustrative dataset
Student	Hours of sleep	Quiz score (of 20)
1	5	11
2	6	14
3	6.5	12
4	7	15
5	7	13
6	7.5	16
7	8	17
8	8	14
9	8.5	18
10	9	17

Worked example — reading r for sleep and quiz score. For these ten students, n (the number of participants in the sample) is 10.

The averages. Mean sleep is 7.25 hours and mean score is 14.7 points.
The coefficient. r (the correlation coefficient) is 0.8457.
Direction. The sign is positive, so the two measures rise together: more sleep tends to go with higher scores.
Strength. 0.8457 is close to +1, so the linear association is strong. The points would sit close to a rising line on a graph.
The sentence. In this sample, students who slept more tended to score higher.

Notice what your sentence does not say. It does not say "sleep raised their scores." Section 4 explains why.

A negative r is as useful as a positive one. If r were close to −1, one measure would fall as the other rose, and the link would still be strong. Only an r near 0 means there is little linear link at all.

In practice (your tools). In Google Sheets or Excel, put hours of sleep in A2:A11 and quiz scores in B2:B11. In an empty cell, type =CORREL(A2:A11,B2:B11). The cell shows 0.8457 when you set four decimals. You will do this yourself in this week's Analysis Workshop.

Misconception check. Students often read an r of 0.8457 as "84.57% of students" or "sleep explains 84.57% of the score." It is tempting because the number looks like a share. But r is not a percentage of anything. It is a measure of how tightly two variables move together on a scale from −1 to +1. The cure: always translate r into words — direction, strength, and "associated with."

4. Explain why correlation is not causation

A hardware store owner notices that on days she sells more umbrellas, the town also reports more fender-benders. Do umbrella sales cause car crashes? No. Rainy weather pushes both numbers up. The rain is a third variable.

That is the first of two reasons a correlation cannot prove a cause.

The third-variable problem is the possibility that an outside factor drives both measured variables in a correlation, so the two move together even though neither causes the other.
The directionality problem is the possibility that the causal arrow in a correlation runs the opposite way from the one a reader assumes, so the second variable may be affecting the first.

Worked example — what else could explain the sleep data? Return to the sleep data, where r (the correlation coefficient) is 0.8457.

A third variable. A student who slept more may also have finished studying earlier. Finishing early could make it easier to sleep and also raise the score. Other candidates: a lighter work schedule, or feeling less anxious about the quiz.
Directionality. Maybe a good day of studying made sleep easier. Then the score's cause, good preparation, came first, and the sleep followed.
The verdict. The design supports "associated with" and nothing stronger.

None of this means sleep does not matter. It means this design cannot tell us. A different design could.

Misconception check. Students often think a strong correlation must mean one thing causes the other. A strong r feels like proof. But strength and cause are separate questions. Umbrella sales and fender-benders could have a strong correlation, and neither causes the other. The cure: after every correlation, name one third variable and one reversed arrow before you say anything else.

Pause and think. Choose one of these and answer it in two or three sentences. (a) Think of a correlation you have heard in the news or on social media. What third variable could explain it? (b) Think of a habit you believe helps you study. What would you need to measure to check your belief fairly?

5. Run an experiment: change one thing and measure the result

Suppose you coach a youth soccer team and want to know whether a new warm-up reduces injuries. If you let players choose the warm-up, the careful players might pick it. Then any drop in injuries could come from the players, not the warm-up. To find a cause, you have to decide who gets which warm-up, and change only that.

That is an experiment.

An experiment is a design in which the researcher manipulates one variable, measures its effect on another, and randomly assigns participants to conditions so that a difference between the groups can be traced to the manipulation.
The independent variable (IV) is, in an experiment, the variable the researcher deliberately manipulates or changes between groups.
The dependent variable (DV) is, in an experiment, the variable the researcher measures to see whether the manipulation had an effect.
The experimental group is the participants in an experiment who receive the manipulation or treatment being tested.
The control group is the participants in an experiment who do not receive the manipulation, giving a baseline for comparison with the experimental group.

Worked example — Loftus and Palmer (1974): the verb and the speed estimate. Elizabeth Loftus and John Palmer showed students short films of car accidents. Then they asked one question: "About how fast were the cars going when they ___ each other?" The blank held a different verb for different students.

Label the variables. The IV is the verb in the question. The DV is the speed estimate in miles per hour.
Experiment 1 (45 students). The table below shows the mean estimate for each verb.
The gap. "Smashed" averaged 40.8 mph and "contacted" averaged 31.8 mph. That is a gap of 40.8 − 31.8 = 9.0 mph, from changing one word.

Mean speed estimate in miles per hour by verb, Loftus and Palmer (1974), Experiment 1, 45 students
Verb in the question	Mean estimate (mph)
smashed	40.8
collided	39.3
bumped	38.1
hit	34.0
contacted	31.8

Experiment 2 (150 students, 50 per condition). A week later, students were asked whether they had seen broken glass. There was none in the film. In the "smashed" group, 16 ⁄ 50 = 32% said yes. In the "hit" group, 7 ⁄ 50 = 14% said yes. In the control group, which was asked no speed question, 6 ⁄ 50 = 12% said yes.
The comparison. That is 32 − 14 = 18 points more for "smashed" than for "hit." Put another way, 32% versus 14% is more than 2 times the rate.

In words: about one in three students who heard "smashed" remembered glass that was never there. The control group shows what happens with no leading verb at all.

What lets this design claim a cause? Random assignment.

Random assignment is placing each participant into a condition of an experiment by chance, so the groups start out alike on average and a difference between them can be traced to the manipulation.

Because participants were randomly assigned to a verb, the verb is the only planned difference between the groups. Careful drivers, distracted viewers, and people who are good with speeds end up spread across every group by chance. So the design lets you say "the wording changed the estimate." That is a causal claim, and the design earns it.

Misconception check. Students often think the IV is "the important variable" or "the one you care about most," and then they swap the labels. It is tempting because the speed estimate feels like the interesting part. But the labels are about roles, not importance. The IV is what the researcher changes. The DV is what the researcher measures. The cure: ask "what did they change?" for the IV and "what did they record?" for the DV.

6. Keep two random procedures apart: sampling and assignment

Imagine a company with 2,000 employees that wants to test a new afternoon break. A manager picks 100 employees by drawing names from the full staff list. Then she flips a coin for each person: heads gets the new break, tails keeps the old schedule. She has used chance twice, for two different jobs.

A population is the whole group of people a researcher wants to draw conclusions about.
A sample is the smaller set of people from a population who actually take part in a study.
Random sampling is choosing participants so that every member of the population has an equal chance of being selected, which lets researchers generalize from the sample to that population.

Worked example — who is in, and who gets what. Return to the manager's study.

Random sampling. Drawing 100 names from all 2,000 employees decides who is studied. It lets her generalize to the whole staff.
Random assignment. The coin flip decides who gets which condition. It lets her claim the new break caused any difference.
Check the numbers. If the coin splits the sample evenly, 50 + 50 = 100 people, with 50 in each condition.

Now change one step and see what you lose. Suppose she skips the name draw and uses only the 100 volunteers from one floor. The coin flip still supports a cause, but only for people like those volunteers. Suppose instead she draws names fairly but lets people choose their break. Now she can generalize, but she cannot claim a cause.

Misconception check. Students often think random sampling and random assignment are the same thing, because both use chance. They do two separate jobs. Random sampling is about who gets into the study, and it supports generalizing. Random assignment is about which condition each person gets, and it supports a cause. The cure: for any study, ask both questions separately — "who was picked?" and "who got what?"

7. Check the measure itself: reliability versus validity

Step on your bathroom scale three times in a row. If it reads the same number each time, it is consistent. Now suppose that number is always five pounds lighter than your real weight. The scale is consistent and also wrong.

Your scale shows the difference between two qualities of any measure.

Reliability is the consistency of a measure: a reliable measure gives the same or a very similar result when it is repeated under the same conditions.
Validity is the accuracy of a measure: a valid measure captures the construct it claims to measure; a measure can be reliable without being valid.

Worked example — a "focus" score. A study app claims to measure focus by counting how many times you open the app each day.

Is it reliable? Probably. If your habits stay the same, the count will be similar from day to day.
Is it valid? Doubtful. Opening the app often might mean you are focused, or it might mean you are distracted and checking your phone.
A better operational definition. Focus could be measured as the number of correct answers on a short task while a timer runs, with distractions controlled.

Misconception check. Students often think a measure that gives the same answer every time must be accurate. Consistency feels like proof. But the five-pounds-light scale is perfectly consistent and always wrong. The cure: check reliability and validity as two separate questions. A measure needs both.

8. Protect the people in the study: research ethics

Imagine you sign up for a study about memory. Halfway through, the researcher tells you to keep going even though you feel upset. Can you leave? Who checked this study before it began? What will happen to your answers?

Those questions are what research ethics answers for you. Psychology learned many of these rules the hard way.

Worked example — Milgram (1963) and why the rules exist. Stanley Milgram told participants they were helping with a study of learning. Each participant was told to give electric shocks to a "learner" for wrong answers, rising to a switch labeled 450 volts. In fact the learner was working with the researcher, and no real shocks were given. Milgram (1963) reported that 26 ⁄ 40 = 65% of participants continued to the 450-volt switch. They believed the shocks were real. Many showed signs of strong stress.

The finding itself belongs to Week 13, where the course studies obedience. This week the study matters for a different reason. Its participants were deceived, put under heavy stress, and pushed to continue. Studies like it are part of why the modern rules exist.

A second piece of history is harsher. From 1932 to 1972, the U.S. Public Health Service followed Black men with syphilis without their informed consent and without offering them available treatment, even after penicillin became widely available. The outcry after this study at Tuskegee was exposed led to the National Research Act of 1974. That law's national commission wrote the Belmont Report's ethical principles, and it led to the modern system of review.

Today, the American Psychological Association's Ethical Principles of Psychologists and Code of Conduct (Standard 8) sets out the protections. Every study with people must provide them.

Informed consent is a participant's voluntary agreement to take part in a study after being told what it involves, its possible risks, and the right to stop at any time.
The right to withdraw means a participant may stop at any point without penalty.
Protection from harm means the study must keep physical and emotional risk as low as possible.
Deception is withholding or misstating the true purpose of a study; it is allowed only when the question cannot be studied another way and is not expected to cause harm, and it must be followed by debriefing.
Debriefing is the explanation given to participants after a study ends, revealing its true purpose and any deception and answering their questions.
Confidentiality is the research rule that information about individual participants is kept private and is not shared in a way that identifies them.
An Institutional Review Board (IRB) is a committee that reviews a proposed study with human participants and must approve it as ethical before the study begins.

Misconception check. Students often think deception is either always forbidden or always fine once people sign up. Neither is right. Deception is allowed only when the question cannot be studied another way and no harm is expected, and it must be followed by debriefing. The cure: when a study uses deception, ask three things. Was there another way? Was harm expected? Were people told the truth afterward?

9. Use the three-question move on a "studies show" headline

Here is a headline you might scroll past on your phone: "Study shows people who eat breakfast do better in school." Before you share it, run three questions.

What design produced it? Did researchers measure breakfast and school results as they happened, or did they assign people to eat breakfast?
What can that design conclude? A correlation earns "associated with." Only an experiment with random assignment earns "causes."
What third variable could explain it? Families with steady morning routines might also have more time for homework.

Worked example — two headlines, one habit. Compare two studies from this chapter.

The sleep data. Design: correlational, since nobody's sleep was changed. Conclusion allowed: in this sample, students who slept more tended to score higher, with an r of 0.8457. Third variable: finishing studying earlier.
The verb study. Design: experiment, since the verb was assigned at random. Conclusion allowed: changing "hit" to "smashed" changed what students reported, from 14% to 32% for broken glass. Third variable: random assignment spread those across the groups, so none is left to explain the gap.

Your habit is the same for any claim about people. It works for a wellness ad, a news story, and a chatbot's confident answer.

Misconception check. Students often think "a study shows" means the question is settled. A headline is short, and it usually drops the design. But one study, of one design, with one sample, can only claim so much. The cure: never repeat a headline until you can answer all three questions about it.

Key terms

Operational definition — The exact, observable procedure a researcher uses to measure a construct, such as measuring memory as percent savings on relearning; a construct without one is not yet measurable. (first used: 1. Run the research loop: turn a fuzzy idea into a measurement)
Replication — Repeating a study, with the same or new participants, to see whether its result appears again; a finding that does not replicate is not kept as settled. (first used: 1. Run the research loop: turn a fuzzy idea into a measurement)
Percent savings — A measure of memory: how much less time or effort relearning material takes than learning it the first time, written as a percentage of the original learning. (first used: 1. Run the research loop: turn a fuzzy idea into a measurement)
Descriptive research — Research that records what people do, say, or report as it happens, through case studies, naturalistic observation, or surveys, without testing what causes it. (first used: 2. Describe behavior: case studies, observation, and surveys)
Case study — A descriptive method that examines one person or a small group in great depth; it can suggest ideas to test but cannot show cause or be assumed to fit everyone. (first used: 2. Describe behavior: case studies, observation, and surveys)
Naturalistic observation — A descriptive method in which a researcher watches and records behavior in its everyday setting without trying to change it. (first used: 2. Describe behavior: case studies, observation, and surveys)
Survey — A descriptive method that asks many people to report their own attitudes, beliefs, or behavior, usually through a questionnaire or an interview. (first used: 2. Describe behavior: case studies, observation, and surveys)
Correlational study — A design that measures two or more variables as they naturally occur and reports how strongly they move together; nothing is manipulated, so it has no independent or dependent variable. (first used: 3. Read a correlation: direction, strength, and the "associated with" sentence)
Correlation coefficient (r) — A number from −1 to +1 that describes a linear association between two measured variables: its sign gives the direction and its size gives the strength, and a value near 0 means little linear association. (first used: 3. Read a correlation: direction, strength, and the "associated with" sentence)
Third-variable problem — The possibility that an outside factor drives both measured variables in a correlation, so the two move together even though neither causes the other. (first used: 4. Explain why correlation is not causation)
Directionality problem — The possibility that the causal arrow in a correlation runs the opposite way from the one a reader assumes, so the second variable may be affecting the first. (first used: 4. Explain why correlation is not causation)
Experiment — A design in which the researcher manipulates one variable, measures its effect on another, and randomly assigns participants to conditions so that a difference between the groups can be traced to the manipulation. (first used: 5. Run an experiment: change one thing and measure the result)
Independent variable (IV) — In an experiment, the variable the researcher deliberately manipulates or changes between groups. (first used: 5. Run an experiment: change one thing and measure the result)
Dependent variable (DV) — In an experiment, the variable the researcher measures to see whether the manipulation had an effect. (first used: 5. Run an experiment: change one thing and measure the result)
Experimental group — The participants in an experiment who receive the manipulation or treatment being tested. (first used: 5. Run an experiment: change one thing and measure the result)
Control group — The participants in an experiment who do not receive the manipulation, giving a baseline for comparison with the experimental group. (first used: 5. Run an experiment: change one thing and measure the result)
Random assignment — Placing each participant into a condition of an experiment by chance, so the groups start out alike on average and a difference between them can be traced to the manipulation. (first used: 5. Run an experiment: change one thing and measure the result)
Population — The whole group of people a researcher wants to draw conclusions about. (first used: 6. Keep two random procedures apart: sampling and assignment)
Sample — The smaller set of people from a population who actually take part in a study. (first used: 6. Keep two random procedures apart: sampling and assignment)
Random sampling — Choosing participants so that every member of the population has an equal chance of being selected, which lets researchers generalize from the sample to that population. (first used: 6. Keep two random procedures apart: sampling and assignment)
Reliability — The consistency of a measure: a reliable measure gives the same or a very similar result when it is repeated under the same conditions. (first used: 7. Check the measure itself: reliability versus validity)
Validity — The accuracy of a measure: a valid measure captures the construct it claims to measure; a measure can be reliable without being valid. (first used: 7. Check the measure itself: reliability versus validity)
Informed consent — A participant's voluntary agreement to take part in a study after being told what it involves, its possible risks, and the right to stop at any time. (first used: 8. Protect the people in the study: research ethics)
Deception — Withholding or misstating the true purpose of a study; it is allowed only when the question cannot be studied another way and is not expected to cause harm, and it must be followed by debriefing. (first used: 8. Protect the people in the study: research ethics)
Debriefing — The explanation given to participants after a study ends, revealing its true purpose and any deception and answering their questions. (first used: 8. Protect the people in the study: research ethics)
Confidentiality — The research rule that information about individual participants is kept private and is not shared in a way that identifies them. (first used: 8. Protect the people in the study: research ethics)
Institutional Review Board (IRB) — A committee that reviews a proposed study with human participants and must approve it as ethical before the study begins. (first used: 8. Protect the people in the study: research ethics)

Summary

An operational definition turns a fuzzy idea into a procedure anyone can repeat.
Descriptive research describes behavior; it cannot show a cause.
A correlation reports direction and strength with r, and earns only "associated with."
A third variable or a reversed arrow can explain any correlation.
An experiment changes the IV, measures the DV, and uses random assignment to earn "causes."
Random sampling supports generalizing; random assignment supports a cause.
A reliable measure is consistent; a valid measure is accurate; you need both.
Every study with people needs informed consent, the right to withdraw, protection from harm, debriefing after any deception, confidentiality, and IRB approval.

Check your understanding

Work these before looking at the answers — they are ungraded, and the answers (with reasoning) are at the bottom of the page.

A researcher spends a month at a public playground and records how often toddlers share toys, without speaking to anyone. Name the design, and say whether it can show why some toddlers share more.
A hypothetical survey of commuters finds an r of −0.42 between minutes spent commuting and a life-satisfaction rating. Describe the direction and the strength in words, and name one third variable.
A gym recruits 60 volunteers and randomly assigns 30 to run on a treadmill with upbeat music and 30 to run without music. It records how many minutes each person runs. Name the IV, the DV, and the control group, and say what random assignment adds.
A fitness watch reports 1,000 steps every time its owner walks exactly 900 steps. Is the watch reliable, valid, both, or neither? Explain.

Answers (worked reasoning — read after attempting)

This is naturalistic observation, a kind of descriptive research. The researcher watched behavior in its everyday setting and changed nothing. It can describe how often sharing happens, but it cannot show why. Older toddlers, siblings at home, or the toys on hand could all play a part, and the design cannot separate them.
The sign is negative, so as commuting time goes up, life satisfaction tends to go down. The size, 0.42, is moderate: well away from 0 but not close to 1. In words: in this sample, longer commutes were associated with lower life satisfaction. A third variable could be income or housing cost: people who can afford to live close to work may also be more satisfied for other reasons. The survey cannot claim the commute causes the drop.
The IV is whether music plays, because that is what the gym changed. The DV is the number of minutes run, because that is what it measured. The control group is the 30 people who ran without music, and 30 + 30 = 60 accounts for everyone. Random assignment spreads fitness, mood, and motivation evenly across both groups by chance, so a difference in minutes can be traced to the music.
The watch is reliable but not valid. It gives the same reading every time, so it is consistent. But it overcounts by 1,000 − 900 = 100 steps on every walk, so it does not measure true steps accurately. This is the same pattern as a scale that always reads five pounds light.

Next up in this module: the Readings & Resources page, the lecture tutorial, the practice set, and the quiz — and next week the neuron and the synapse, the brain's signaling mechanism, show where behavior begins in the body.

Built with The Course Maker · samples.thecoursemaker.com