Lecture 12 Flashcards Preview

Stats Final > Lecture 12 > Flashcards

Flashcards in Lecture 12 Deck (17):

Categorical Variable

Nominal--categories the sample distinctly falls into


Continuous Variable

Interval--a variable that can be measured to any level of precision


Dependent Variable

Outcome Variable
A Variable whose measure is based on the manipulation of the predictor (IV) variable


Independent Variable

Predictor Variable
A variable that is manipulated in order to make changes on the outcome (DV) variable



Looking at two samples, in a group or in two points in time.


Standard Error of difference between means



When would you conduct a one-sample t-test?

*Used to compare the sample mean of a variable (DV) with a "test value"

Example: Determine if depression is more/less apparent in teenage boys who play sports, using a standardized test called the KUDI.
You test 30 subjects, and compare their outcome sample mean to the standardized score set for teenage boys.


When would you conduct a paired-samples t-test?

*Used to compare means of a single sample in a longitudinal design with only two time points (e.g., pre and post test).

*Can be used to compare the means of two variables (e.g. depression and quality of life).


When would you conduct an independent-samples t-test?

*Used to compare the means of two independent samples (a subject cannot be in both groups) on a given variable


Requirements of IV and DV in an independent-sample t-test

*One categorical/nominal IV, with two levels or groups

*One continuous/interval/ratio DV


What is the null hypothesis for all the types of sampled t-tests?

The mean difference between the two comparison groups = 0


What is the null hyp for Levene's test of the equality (of error) variances?

It evaluates the variance between the groups to ensure the assumption of homogeneity of variances.

Ho = Xa = Xb


What if the outcome of the null hyp for Levene's test is not significant?

If it is not significant, then we assume that we have equal variances.

**If p > .05, accept the null


What if the outcome of the null hyp for Levene's test IS significant?

If it is significant, we MAY NOT assume that we have equal variances.

**If p


If given a 95% confidence interval around two group means for an independent-samples t test, be able to
test the null hypothesis that the mean difference between two group means is zero (or that the two group means are equal).

1. Examine Levene's test for equality of variances
**SPSS finds it for you, along with the F-value.
** Find the p- value (also 1 - F)
**Compare p-value with .05...is is significant or not?

2. Depending on if it is significant or not, use the corresponding data in the SPSS output chart.

3. Looking at the CI values, determine if 0 would fall in that range.
**If 0 is present, you fail to reject the null, because the difference between the two means is not significantly different from 0


Comparing the observed t to the critical t

observed t > critical = significant/reject null


What does the t value comparison tell us

It determines if the difference in means is statistically different from zero