Week 7 - T-Tests & Chi-square Flashcards
(54 cards)
Covariance
measure of how much those diffs from one variable occur togther with the diffs of another.
to determining the association between two variables we look for this
Variance
meaure of the diffs of your sample measure from from the sample mean
How spread out are the scores in your sample
scatter plot is used?
to see the associations/corrolation of two variences
correlation
a test for how much two variables co-vary togther on a standardized scale
how related/associated are two varibles
-1 - 0 - +1
Pearson’s Correlation Assumptions
planning:
- people’s measurements are independent to others
- the variables are normally distributed (larger sample is good)
After:
- your data do not have outliers or non-linear patters
for what processes are P-values used?
Deductive Processes
only exists in quantitative research
D = Theory -> hypothesis -> Data
Hypothese and P-values
H0: there is nothing
H1: there is something
p tests the liklyhood of something being diff.
H1: something
- diff between groups
- association between two variables
- experimental change
what is the p-value
the probability that we would find a diff/ass of that size (or bigger), if there was actually no diff/ass
p is an idicator of a false positive
low = low chance and supports H1
high = high chnace and supports H0
statistically significant
.05 =/<
(5%)
= or < .05 means we reject H0
Direction
-/+
positive = increase in one v relates to an increase in the other v (/)
negative = increase relates to a decease()
Strength
r
0.1 = weak/small
0.3 = moderate
0.5 or > = large/strong
Significance
if p-value .05 or < than there is an ass
what does a standard error look like
big SE = overlap between means
small SE = no overlap between means
diff between Std./ variance and Standard Error of Mean
Std = spread of sample
SEoM = spread of estimate
statistical significance and SE
if error bars a far apart = sig
what is an Idenpendent samples t-test?
a test of whether the means of two independent groups are diff.
Idenpendent samples t-test variables
Categorical
Ordinal or Continuous
Independent samples Student’s t-test Assumptions
before:
- each measurement in the sample is independent
- the varibles are normally dist (>30 samples)
after:
- Variances of the two groups are relatively equal
Cohen’s d
.2 = small
.5 = Medium
.8 = large
what test for tow means from one sample
paired-sample t-test
what test for comparing means from three or more groups
ANOVA
what test for one mean from one group compared to a fixed value?
One sample t-test
Chi-Square
a test of the relationship or ass between two categorical varibles
