Exam 2 Study Guide

Draw an interaction between two variables in an experiment.*

Pic ->

How many hypotheses would you need to at least consider for a factorial design with two IVs and what would each look like?*

- 2 x 2, 3 hypotheses - 2 x 3, 3 hypotheses (2 main effects) (1 interaction)

What are the advantages and disadvantages of correlational designs?*

- ADV. Correlational research allows researchers to collect much more data than experiments. it opens up a great deal of further research to other scholars. It allows researchers to determine the strength and direction of a relationship so that later studies can narrow the findings down and, if possible, determine causation experimentally. - DISADV. Correlation research only uncovers a relationship; it cannot provide a conclusive reason for why there's a relationship. a third, unknown variable might be causing both.

what is a correlational research, bivariate correlation, when should it be used, and what kinds of conclusions can we draw from this design?*

- a study with all measured variables is correlational. - bivariate correlation, is an association that involves exactly two variables. - larger effect sizes allow more accurate predictions. - larger effect size are usually more important - correlation coefficient or ratio is closer to one, means a stronger relationship. - able to see outliers (effect scores) and identify mistakes in calculations.

What is the difference between confounding variables and (non-confounding) thirdvariables? How do you know if you have anon-confounding third variable or a confounding variable present? How does the presence of each affect the results/conclusions of our study?

- confound - type of 3rd variable, systematically different from IV. several possible alternate explanations, potential threats to internal validity. - If there is a clinically meaningful relationship between an the variable and the risk factor and between the variable and the outcome (regardless of whether that relationship reaches statistical significance), the variable is regarded as a confounder. - If you have accounted for any potential confounders, you can thus conclude that the difference in the independent variable must be the cause of the variation in the dependent variable. - A confounding variable is an "extra" variable that you didn't account for. They can ruin an experiment and give you useless results. They can suggest there is correlation when in fact there isn't. They can even introduce bias.

What are the four validities and how can we assess experiments in terms of each?*

- construct, how well were the variables measured and manipulated? - manipulation check, an extra dependent variable that researchers can insert into an experiment to convince them that their experimental manipulation worked. - pilot study, simple study using a separate group of participants that is completed before conducting the study primary interest. - internal, are there alternative explanations for the results? - no confounds or third variables. - random selection. control for order of effects by counterbalancing. - external, to whom or what can the casual claim generalize? - random sampling, random gathering a sample from population. - random assignment, random assigning each participant in a sample into one experimental group of interest. - statistical, how well do the data support the casual claim? - r, evaluates the effect size of an association. - d, represent how far apart two experimental groups are on scores within each group.

What is experimental research and what is its purpose? How does experimental designs meet the three criteria for establishing causality?*

- experimental research - researcher manipulated at least one variable and measured the other. - purpose, The purpose of the experimental research strategy is to establish the existence of a cause- and- effect relationship between two variables. - covariance, provides the comparison group by answering the question "compared to what?" manipulating the IV is necessary. (control group, treatment group, placebo group.) - temporal precedence, causal variable did come before the outcome or DV. by manipulating the IV, the experimenter virtually ensures that the cause comes before the effect. - internal validity, ensures that the casual variable (active ingredient) and not other factors are responsible for the change in the outcome.

How do we evaluate experimental designs in terms of external validity? Be able to compare and contrast the importance of internal and external validity.*

- external validity, ask how the experimenters recruited their participants. (random sampling) - external validity applies to types of situations to which an experiment might generalize. - internal validity, ask about random assignment. - well designed experiment establishes internal validity. (most important validities to interrogate when you encounter causal claims.) - experimenters usually prioritize internal validity.

What is a factorial design and how do we determine the notation for a factorial design? How do we determine the number of cells or conditions within a factorial design?*

- factorial design - study in which there are two or more IV or factors. - notation is (# of # IV, values of levels.)(how many?) - _ x _ = cells

What is counterbalancing and why might we use counterbalancing?*

- in a repeated measures experiment, presenting the levels of the IV to participants in different sequences to control for order effects. - purpose, within groups design, they have to split participants into groups: each group receives one of the condition sequences. random assignment to determine the sequence. - full counterbalance, all possible condition orders are represented. - partial counterbalance, some of the possible conditions in a randomized order for every subject.

Why is it so important to accept and appreciate null results? Be able to couch your answer to this question within broad themes of this class.*

- instead of the case that the IV really does not affect the DV, in other words, the experiment gave an accurate result, showing that the manipulation the researchers used did not cause a change in the DV. - by obtaining the null, he state our theory is incorrect. - maybe study was not designed or conducted carefully. - maybe some obscuring factor in the study presented the researchers from detecting the true difference.

How can we interpret null effects? How can we use statistics to support our interpretations of null effects?*

- interpreting null effects, no significant covariance between IV and DV - ceiling effect, all scores are squeezed together at the high end. - floor effect, all scores cluster at the low end. - shows there is not enough between-group differences to produce a significant outcome.

How does each option relate to the three rules for establishing causal explanations?*

- longitudinal design, provides evidence for temporal precedence by measuring the same variables in the same people at several points in time. - cross sectional test whether two variables are correlated (covariance) - autocorrelations (covariance) because it determines the correlation of one variable with itself, measured on two different occasions. (longitudinal) - cross lag correlation helps with temporal precedence and internal validity.

What multivariate options can we employ to strengthen the conclusions we draw from correlational data?*

- longitudinal design, provides evidence for temporal precedence by measuring the same variables in the same people at several points in time. - cross sectional, test whether two variables measures at the same point in time, are correlated. - autocorrelations, determine the correlation of one variable with itself, measured on two different occasions. - cross lag correlations, show whether earlier measure of one variable is associated with later measure of the other variable. Directly comparing a to b or b to a.

What is a main effect? What is an interaction? How do we write hypotheses for each?*

- main effect - overall effect of one IV on the DV, averaging over the levels of the other IV. (effect expect on one IV on DV, ignoring all other variables) "Main Effect: Instructions written in red color will be perceived as more difficult than instructions written in green. Main Effect: Instructions written in cursive will be perceived as more difficult than instruction written in normal type face." - interaction - differences in the levels of one IV changes, depending on the level of the other IV, difference in differences. (joint effect expected on IV's and DV) "Spreading, Instructions written in cursive will be perceived as more difficult than normal typeface, and the difference will be greater for instructions in red font than green font."

How do you interpret main effects? How do you interpret a two-way interaction? Be able to do each of these using descriptive statistics and/or a graph to guide your interpretation.*

- main effects interpretation, an overall effect of one IV on the DV averaging over the levels of the other IV. The overall effect of one iV at a time. - two-way interaction interpretation, the third result, effect is the difference in differences.

What is the difference between a mediating variable, a moderating variable, and a third variable?*

- mediator - a variable that helps explain the relationship between two other variables. - moderator - variable that depending on its level, changes the relationship between two other variables. (A has an relationship with C who has a relationship with B.) - third variable - existence of a plausible alternative explanation for the association between two variables.

what is a multivariate correlation, when should it be used, and what kinds of conclusions can we draw from this design?*

- multivariate correlation, involving more than two measured variables. - conclusions, positive beta, indicates a positive relationship. vise versa. beta that is 0, is not significant. - CORRELATION coefficients describe the strength and direction of an association between variables. A Pearson correlation is a measure of a linear association between 2 normally distributed random variables. A Spearman rank correlation describes the monotonic relationship between 2 variables.

How is our interpretation of β similar to and different from our interpretation of B? What does it mean to say that β represents a "partial correlation?"*

- positive beta indicates positive relationship (r) between predictor variable and criterion variable. negative is a negative relationship. when beta is 0, that represents no significance or relationship. (note the direction and strength of a relationship. higher the beta, the stronger the relationship. vise versa) (standardized units) benefit is how many standard deviations away from the mean, make a direct comparison. - b (estimate) represents an unstandardized regression coefficient. similar to beta in that positive or negative denotes a positive or negative association. Unlike two betas, we cannot compare two b values within the same table. (dollars, centimeters, or inches) - beta represents a partial correlation because we can compare standard deviations with one another.

What is a spreading interaction? A crossover interaction? Be able to identify, interpret, and draw each.*

- spreading interaction - (only when) strong on one level compared to the other level. "my dog says sit, but only when I am holding a treat." - crossover interaction - (it depends) "it depends on the temp at which the ice cream or pancakes are served."

How do we interpret correlational findings and evaluate their validity?*

- statistically significant (p) is the probability associated with the result is very small, less than 5%, the result is very unlikely to have come from a zero association population, the correlation is considered significant. - nonsignificant, we cannot rule out the probability that the result came from a population in which the association is zero. - mean, average - t test, the test whether between mean (group averages) is statistically significant. - construct variable, match between operational definition and measure.only measured variable constitutes entire study. face, content, criterion related, convergent, discriminant. - internal validity (sticking point) extent to draw causal conclusions. (strictly X related to Y.)

What are the three rules for establishing causal explanations?*

1. covariance of cause and effect - results must show a correlation, or association between the cause variable and effect variable. (relationship) 2. temporal precedence - time ordering, the cause variable must precede the effect variable, must come first in time. 3. Internal validity - no plausible alternative explanations for the relationship between two variables. (no 3rd variables)

What are the basic features of an independent-groups design? What types of independent-groups designs exist? What are the tradeoffs between each?*

ID groups - different groups of participants are placed into different levels of the IV. also called between groups. - no connections between groups at all. - Between-subjects experiments have the advantage of being conceptually simpler and requiring less testing time per participant. They also avoid carryover effects without the need for counterbalancing.

What threats to validity exist for experimental designs? Be able to identify and describe—in your own words—each threat. How can we protect against each threat? Which threats apply more to independent-groups designs? Within-groups designs?*

chapter 11 quizlet.

What are the basic features of a within-groups design? What are the different types of within-groups? What are the advantages and disadvantages of this type of design.*

within groups - one group of participants, each person is presented with all levels of IV. - groups are linked in someway. (same people, repeated measures) naturally occurring pairs (siblings, parents/child.) matched pairs (researcher matches pairs) - benefit, RM are smaller number of participants (30 min) so you do not have to worry about individual differences. (same people are natural, important by reducing link of individual differences increasing statistical power.) - disadvantages, in RM, participants figure out what's going on and change their behavior. or causing fatigue and practice effects. - Within-subjects experiments have the advantage of controlling extraneous participant variables, which generally reduces noise in the data and makes it easier to detect a relationship between the independent and dependent variables.