Research Methods Exam 2

How can you test an association claim?

Through correlational research using graphs and simple statistics

What is a bivariate correlational research?

An association that involves two variables. The research is to find out the relationship between the two variables.

What is a Pearson correlation coefficient?

The Pearson correlation coefficient is r 

What’s the difference between a positive and negative Pearson correlation coefficient?

A positive correlation means that the two variables are moving upwards, whilst a negative correlation means the two variables are moving downwards

How can you tell the direction of the Pearson correlation coefficient?

The positive means it is moving up and the negative sign means it is moving downward

How can you tell the magnitude/strength of the Pearson correlation coefficient?

0 means there is no relationship, the closer to positive or negative 1 the stronger the relationship between the two variables

What can you conclude from the results of correlational research?

Can conclude if the two variables have a relationship

What is multivariate vs. bivariate correlational research?

Multivariate correlational research is a study designed to test an association involving more than two measured variables.

A bivariate correlational study is exactly two measured variables 

What are the three criteria for casual claims?

Covariance, temporal variance, and internal validity

What are two types of multivariate correlational designs?

Cross-sectional correlation and cross-lag correlation

Cross-sectional correlational

Longitudinal design correlation between two variables that are measured at the same time

Cross-lag correlation

Longitudinal design, a correlation between an earlier measure of one variable and later measure of another variable

What are longitudinal designs?

Studies in which the same variables are measured in the same people at different points in time

What can longitudinal studies provide us?

Used in developmental psychology to study changes in a trait or an ability as a person grows older

How do longitudinal designs help with temporal precedence?

Each variable is measured at clearly different points in time, they know which one came first

What is multiple regression?

A statistical technique that computes the relationship between a predictor variable and a criterion variable, controlling for other predictable variables 

How does multiple regression analysis help with the third-variable problem?

It holds the third variable at a constant level (statistically or experimentally) while investigating the association between the two other variables

What does it mean to statistically control for a variable?

Testing a third variable with multiple regression means identifying subgroups, or proportions of variability, and asking whether the bivariate relationship holds on all levels

What is moderation?

Can change the relationship between the two other variables (making it either more or less intense)

What kind of questions does moderation answer?

Are there certain groups or situations for which two variables are more strongly related? Who is the most vulnerable?

What is mediation?

A variable helps explain the relationship between two other variables

What kind of question does mediation answer?

Why are the two variables linked?

What is a scientific experiment?

Researchers manipulated at least one variable and measured another

What goes into the design of a simple experiment?

Manipulated and Measured Variables

What is the IV?

A variable that is manipulated, can explain variance in criterion variable in multiple-regression

What is a DV?

A variable that is measured. In multiple regression the single outcome, or criterion variable the researchers are most interested in understanding or predicting

What does “making all else equal” refer to?

Considering two or more situations where only variable difference is the factor being examined, potentially relevant factors are held constant or are considered to be the same

What is a between-participants design?

Different groups of participants experience different levels of the independent variable

What is a within-participants design?

All participants experience all levels of the independent variable

Advantages of between-participants designs

Usually a simpler design order effects are not a concern

Disadvantages of between-participants designs

Selection bias need a lot of participants for the control group

Advantages of within-participants design

Everyone is in their own control group, fewer participants

Disadvantages of within-participants design

Order effects may not be possible or practical

Posttest-only design

A research study where data is collected only after a treatment or intervention has been applied, without any prior measurement of the outcome

Pretest design

Experiment using independent groups design, participants tested on key DV twice: once before and once after exposure to IV

Repeated measures design

Experiment using within-participants design. Participants are exposed to DV more than once after exposure to each level of IV

Concurrent measures design

Experiment using within groups design, participants are exposed to all levels of an IV at roughly the same time, single attitudinal or behavioral preference is DV

Covariance

Variables need to be related or associated with one another

Temporal Variance

Order of when the variables occurred, which one was first and so worth

Internal validity

Extent to which an experiment accurately reflect a cause-and-effect relationship between variables, minimizing the possibility of alternative explanations for the observed effects

Question that covariance answers

Do the results show that the casual variable is related to the outcome variable? Are distinct levels of the IV associated with different levels of the dependent variable?

Question temporal presence answers

Does the study design ensure that the casual variable comes before the outcome variable in time?

Question that internal validity answers

Does the study design rule out alternative explanations for the results

What is a confounding variable?

A potential third explanation for a research finding. A factor that varies systematically with the independent variable

What are order effects, and how are they problematic?

Found in within-group designs, threat to internal validity. Exposure to 1 condition changes participants responses to a later condition.

Identify the three types of order effects

Carryover effects, practice effects, and fatigue

Carryover effects

IV level 1 influences IV level 2.

Practice effects

IV level 1 taught participants how to complete IV level 2 better

Fatigue

IV level 1 exhausted participants, who are now performing their best on IV level 2

What are selection effects and how are they problematic?

Threat to internal validity occurs in an independent groups design when kinds of participants at 1 level of IV are systematically different from those at another level

What are design effects, and how are they problematic?

Quantifies the extent to which the expected sampling error in a survey deviates from the sampling error. Inflates standard errors, leading to inaccurate conclusions about the statistical significance of relationships in complex samples.

Weighting

Used to account for unequal selection probabilities or to adjust for non-response and non-coverage. It can increase the variance of a statistic. Do not perfectly reflect the true population distribution

Clustering

Grouping sampling units into clusters can also increase variance, as individuals within a cluster tend to be more similar than a random sample.

Stratification

Dividing the population into subgroups. Sampling independently within each stratum can reduce variance if relatively homogeneous. If there's a difference in variability between strata, stratification can increase the variance.

What is random assignment, and what solutions can it provide?

Each participant has an equal chance of being assigned to each of the treatment conditions. Used to control environmental variables. Helps prevent non-random, standard, or systematic error

What solutions can random assignment not provide?

It cannot reflect our population on every relevant dimension, therefore not addressing external validity

What is counterbalancing, and what solutions can it provide?

Different subsets of participants complete conditions in different orders. It can prevent order effects

What is a complex experiment? How is it different from a simple experiment?

When a researcher manipulates the IV and subsequently measures the DV. It has more than one measured or manipulated variable

What can a complex experiment give us? Why should we conduct an experiment with two independent variables?

It supports casual change, and the only thing that varies systematically is the IV. Allows researchers to explore more complex relationships between variables and uncover potential interactions that may have been missed when only examining one variable at a time

What is an interaction? What is another name for an interaction question?

The effect of one independent variable depends on the level of another independent variable (moderation). Another name is interplay or interactivity questions

What is a factorial design?

Study with 2 or more IV or factors

Construct validity

Indication of how well a variable was measured or manipulated in a study

Reliability

Same answers every time. It can prove validity, but validity can’t prove reliability 

Interrater reliability

Two or more independent observers will come up with consistent (or similar) findings

Test-retest reliability

Measures the stability of scores of a stable construct obtained from the same person on two or more separate occasions

Internal consistency

Design and behavior of an application remaining largely the same within screens and features 

Face validity

Face value, does it seem to measure what we are investigating?

Content validity

Representative of our measurements 

Convergent validity

Scores on measures of similar construct are related

Divergent validity

Demonstrates a test’s ability to distinguish between different unrelated constructs

Concurrent criterion-related validity

Comparting behaviors to self-reports

Predictive criterion-related validity

How well a test is predetermined

Statistical validity

How well the numbers measured represent the numbers we think we measured 

External validity

Indication how well results of a study generalize to, or represent individuals or contexts besides those in the study itself

Ecological validity

Extent to which the tasks and manipulations of a study are similar to real-world contexts

Internal validity

Casual relationship between two variables is genuine