Research Methods Exam 1

The steps of the Theory-data cycle

1. Theory

2. Reseach Question

3. Research design to test a specific hypothesis

4. Preregiestered

5. Data can lead to..

   1. Supporting- goes back to step 1

   2. Non-supporting- goes back to step 3

Why don’t we say that a study “proves” a theory is true?

Research provides evidence and support of a theory, but “proving” means that the theory is correct 100% of the time, and good theories are falsifiable

What does it mean for a hypothesis to be falsifiable?

When a hypothesis is falsifiable, it means that it can be tested and possibly shown to be a false statement

What is basic research?

Enhance the general body of knowledge about a particular topic (how and why something works)

What is applied research

Solve practical problems; findings directly applied

Why is journalists’ coverage of science often misleading?

Coverage is often misleading due to the telephone effect; other sources hear it from others and others which then may get the story twisted. (distorted)

How can we reduce the chance of being misled by popular media coverage of science?

By using sources that are peer-reviewed or scholarly journals instead of large feed news articles that are most likely not checked for accuracy

What is empiricism?

The idea that we rely on systematic observation and evidence to learn rather than opinions or personal experiences

How/why is empiricism better than other methods of getting at the truth?

It is..

Measurable

Can be replicated

Reduces biased behaviors

Why is it problematic that personal experiences often have no comparison groups?

Because there is no way to prove that something actually caused the outcome they are stating

What are confounds?

Other explanations for the result

How does empirical research deal with the problem of confounds?

By systematically comparing conditions and factors that could affect results

Group 1: uses new method

Group 2: uses original method

What does it mean to say that research results are probabilistic?

Findings are not expected to explain all cases all the time

availability heuristic

When things are easier to recall, we tend to overestimate their prevalence

confirmation bias

We only seek out and accept info that confirms beliefs

Bias-blind spot

We think we are less biased than others

When might it make sense to trust an authority figure on a subject, and when might it not?

TRUST: when it is proven with studies that you can find online and know others agree with that authority figure

DON'T: Based on just their opinion and because they are just higher authority than you

Variables

Something that can have different values or levels across people or situations

Stress level, amount of sleep, test score

Constants

Something that remains the same for everyone in the study

10-minute testing period for each participant in a study

Measured variables

Observed and recorded without manipulation

Hours of sleep participants normally get

Manipulated variables

Controlled

Sleep 4 or 8 hours

Conceptual variables

An abstract idea or concept that researchers want to study

Happiness, stress, depression, motivation

Operational definitions

A definition that clearly and precisely indicates how a concept will be measured or identified

HOW you're going to study and measure it

Score on a 20-item stress questionnaire

Frequency Claim

Describes a particular level or degree of a SINGLE variable ( HOW OFTEN or HOW COMMON something is)

“60% of college students report feeling stressed”

DESCRIPTIVE

Association Claim

One level of a variable is likely to be associated with a particular level of another variable (2 VARIABLES ARE RELATED OR CONNECTED)

“People who get more sleep tend to have better grades”

CORRELATIONAL

Causal Claim

One variable causes changes in the level of another variable

“Caffeine causes increased alertness”

EXPERIMENT

Be able to tell the difference between a correlational study and an experiment.

Correlational = Measure Experiment = Manipulate

Correlational: measure 2 or more variables and see if they are related

Experiment: Manipulate one variable and measure the effect on another variable

Four types of questions

Open-ended

Force-choice

Likert Scale

Semantic scale

PROS and CONS of open-ended questions

PROS

More detailed

Unexpected answers to help with research

Good for exploring deeply into a topic

CONS

Takes longer to analyze

Hard to compare and contrast

Have to code and classify responses

PROS and CONS of Forced-choice questions

PROS

Quick and easy to answer

Easy to compare and contrast

More standardized data

CONS

May not include the participant's actual answer

Forces people to an option that may not fit their response

Options can influence answers

Be able to describe some of the strengths and weaknesses of self-report data.

STRENGTHS

Easy to collect from many people

Measures difficult things that are hard to observe

Inexpensive and efficient

People report their own experiences researchers wouldn’t know otherwise

WEAKNESSES

May not remember accurately

May give socially acceptable answers

Misreport of behavior

Acquiescence bias

Know different response sets and how to try to prevent or detect these problems.

A tendency for people to answer survey questions in a particular way regardless of the actual content of the questions

Acquiescence bias (tendency to agree with statements regardless of what you think)

Fence-sitting (Moderate, being in between two answers)

What are other tendencies and shortcomings that participants bring into the situation? Be able to identify these as well as ways that researchers minimize their impact on the accuracy of survey responses.

Recall problems- not remembering past events accurately

Social desirability

Lack of knowledge

Carelessness

Misunderstanding

What is a bivariate correlation?

Associations that involve exactly two variables

R, ranges from -1 to 1

Both are measured not manipulated

Be able to correctly interpret a correlation using appropriate language.

Positive correlation- increase, increase

Negative correlation- increase, decrease

Near 0 - no relationship between variables (Straight horizontal line)

Know the strengths associated with correlation

r = +.80 → strong positive association

r = +.30 → weak/moderate positive association

r = −.70 → strong negative association

r = −.05 → very weak/no linear association

Outliers

An extreme score that stands out from the rest on the scatterplot

Correlation is sensitive to extreme scores, one outlier can greatly change r

Range Restriction

Happens when your sample doesn’t include the full range of scores on a variable

When there isn’t much variation, the correlation will most likely appear weaker than is would in a wider-ranged sample

Non-linear data

A correlation measures a linear relationship

If two variables have a strong relationship but the pattern is curved/non-linear, r may be close to 0 evem though the variables are clearly related

What are three criteria needed to make a causal claim? If a research study was described, you should be able to evaluate and explain whether the study meets the three criteria or not.

Covariance

Two variables must be related

When X changes, Y changes

Temporal precedence

The cause must happen before the effect

X happens first; Y happens second

Internal validity

There must be no reasonable alternative explanation for the relationship

Researchers need to rule out confounding variables (the third variable that could explain the relationship between the original 2 variables)

What is a spurious association

A relationship between two variables that appears to exist, but is actually explained by a third variable

TEST HELP: When given X-Y, think, “What Z could be related to BOTH X and Y”

What is a mediator?

Intervening variable that links independent to dependent variable

Moderator

Third variable that changes the relationship between the IV and DV