Quantitative Research Final

random assignment

ensures that each participant has an equal chance of being placed in a condition

random sample

ensures that each member of the population has an equal chance of being chosen for a study

are random sampling and random assignment the same thing?

No, if you don’t have random sample, you can use random sampling to still be considered a simple random sample

validity

concerned with accuracy, how well researchers measure what they intend to measure… is argued for conceptually

reliability

concerned with consistency, measuring something in a consistent and stable manner… is assessed numerically

internal validity

accuracy of conclusions drawn from a particular study

external validity

the generalizability of the findings from a research study

one tailed hypothesis

predicts a significant difference in one direction (ex. dogs are more intelligent than cats… our projects)

two tailed hypothesis

predicts a significant difference without indicating the direction (ex. there will be a difference between dogs and cats in terms of intelligence)

null hypothesis

makes a prediction about NO difference or NO relationships

alternative hypothesis

makes a prediction about a difference or a relationship

hypothesis of difference

predicts a difference (causation)

hypothesis of relationship

predicts a relationship (correlation)

variable level: nominal

categories (ex. types of sports)

variable level: ordinal

ranked but don’t know difference (ex. relationship status 0-6 months, 6-12 months, 12+ months)

variable level: interval

ranked and know magnitude of difference between categories (ex. likert scale)

variable level: ratio

categories are ranked, know difference, and has absolute zero point (ex. anything that can be counted and be zero, exam scores, temperature, age)

If P is greater than .05 there is…

no difference or no relationship

if P is less than .05 there is…

a difference or a relationship

what is the purpose of a t test?

to determine if 2 groups means significantly differ on an interval/ratio level

what is the first question you ask when interpreting a t test output file?

is there a difference between the variances (is data spread out around its mean)

when answering the first question of a t test what do we want to find?

we want no difference in the variances so we can assume equality (P> .05, data is spread out evenly around mean)

where to find answer to question 1 of t test

look at sig under levene’s test for equality of variances

what is the second question when interpreting a t test output file?

is there a difference between the means

what do we want to find when answering question 2 of t test

we want to find a difference P<.05 (statistically significant difference between the two variables)

how to find answer to question 2 of t test?

look at box labeled t-test for equality of means and find sig 2-tailed column

what is the third question when interpreting t test output file?

what is the effect size?

what do we want to find when answering question 3 of t test?

bigger is better but anything will do (cohens d)

what are the sizes of cohens d?

small=.2, medium= .5, large=.8

how to report t test information?

The study revealed that IV1 (M=mean) are more/less DV than IV2 (M=mean), t(df)= t#, P sentence…, cohen’s d= #. Thus the hypothesis was supported/not supported.

what is the purpose of an ANOVA test

to determine if 3 or more groups means significantly differ an an interval/ratio level

what is the first question you ask yourself when interpreting an ANOVA output file?

is there a difference between the means of the IVs (write out the three levels)

what do we want to find when answering question 1 of ANOVA?

we want to find a difference between the means P<.05

how to find answer to question 1 ANOVA

look at the Tests Between Subjects Effects box, find sig column, use IV row

what is second question when interpreting ANOVA and what do you want to find?

what is the effect size… bigger is better but anything with do (partial eta squared)

ANOVA levels of partial eta squared

small=.01, medium=.06, large=.14 (look at partial eta squared box for IV column

what is the third question when interpreting an ANOVA output and what do we want to find?

where does the difference lie… we want to be in line with our hypothesis

how to find answer to question 3 ANOVA

look at multiple comparisons box in sig column for any p-value less than .05… if you find one see which IV levels are involved then check descriptives statistics box to see which mean is higher to find answer

How to report ANOVA findings?

This study revealed that IV1 (M=mean#) are more/less DV than IV2 (M=mean#), F(df#, error#) = F#, p sentence>/<, partial eta2= eta#. Thus, the hypothesis was partially supported.

what is the goal of a correlation test

to determine if there is a relationship between two interval/ratio variables

what is the first question you ask when interpreting a Correlation output and what do you want to find?

is there a relationship, we want a relationship (p<.05)

how to find answer to question 1 Correlation?

look at sig row in the box that crosses both variables

what is the second question for correlation output and what do we want to find

what is the direction, we want it to be in line with the hypothesis

how to find answer to question 2 correlation

look at pearson correlation # to see if it is positive or negative

what is the third question when interpreting a correlation output?

what is the strength -1 — 1 continuum

what do we want to find when answering question 3 of correlation?

stronger is better but anything will do (closer to -1 or 1 than 0 then strong in that direction, if closer to 0 than weak relationship)

how to report correlation findings?

This study revealed that as one variable increases/decreases, their other variable increases/decreases, r(N-variable#)= pearson’s correlation#, p sentence.

research validity

accuracy of project as a whole, what you concluded, how accurate it was

measurement validity

how accurate researchers measure what they intend to measure

content validity

do the items properly sample the entire topic/what we intent to measure

criterion related validity

does the technique used match results of previous different techniques (use two different techniques then match scores)

construct validity

does the instrument measure what it is supposed to? (ex. give old validated likert scale and new one to see if scores are related)

ways to manipulate the IV

experimental vs control group, written materials, audio/visual materials, confederates (people in study), and hypothetical scenarios/role playing

4 features of an experimental design

1. random assignment 2. manipulation of the IV 3. measurement of DV 4. control

things that affect control

threshold effect, experimenter effect, hawthorne effect, extraneous variables

threshold effect: affects control

idea that manipulation of IV needs to reach a certain threshold before it changes the DV (ex. video games are violent but only after 4 hours of play)

experimenter effects: affects control

something related to the experimenter impacts the results of the study

hawthorne effect: affects control

doing an observational study and the participants figure out they are being observed and behave differently

extraneous variables: affects control

something unrelated to the study causes a change in the DV

historical flaw: threat to validity

something unrelated to study is happening in the world and causes participants to behave differently (ex. study on alcohol but student dies from alcohol)

maturation: threat to validity

idea that participants matured naturally and not as a result of manipulation of IV

testing flaw: threat to validity

problem when you give measurement more than once and participants become test wise

selection threat to validity

participants participate in study for nefarious reasons (ex. $100 reward so don’t take study seriously)

attrition threat to validity

when people drop out of your study because it is too long, too personal, etc (happens in longitudinal study)