Exam 3 (ch 9-12) Flashcards
the assumed or claimed mean of the population
μo
the estimate of an unknown population mean
μ1
with Null Hypothesis Testing we are
comparing
a proposition based on limited evidence
hypothesis
a proposition of no difference
null hypothesis
also called the hypothesis of equality
null hypothesis
a statement of difference
alternate hypothesis
only one variable is sampled
one sample design
we are comparing our single sample against a known population parameter
one sample design
the process that produces abilities that are accurate where the null hypothesis is true
null hypothesis significance test
T/F in NHST we always assume the alternate hypothesis is true
false
one tail vs two tail significance test: does bud light contain more than 4.2% abv
one tail
is the abv of budlight different from the companies claim of 4.2%?
two tail
AH will entail a direction to the difference (greater than or less than μo)
one tail hypothesis
AH will not entail a direction, just a general difference
two tail hypothesis test
T/F: NHST allows one to provide evidence for the alternate hypothesis but not for the null hypothesis
true
the sampling distribution is most often a
t distribution
probability is chosen as the criterion for rejecting the null hypothesis
significance level
probability of a type 1 error
alpha
difference (between means) so large that chance is not a plausible explanation for the difference
statistically significant
the area on the distribution that is beyond the significance level; “reject the null here”
rejection zone
the number from the sampling distribution that determines whether the null is neglected
critical value
corresponds to the significance level
critical value
rejection of a null hypothesis that is true
type 1 error
failure to reject a null hypothesis that actually is false
type 2
probability of a type 1 error
alpha
probability of a type 2 error
beta
having a large sample size increases or decreases power
increases
statistical test of the hypothesis that a sample mean came from a population with a population mean
one sample t test
t/f: one sample t test does not have access to parameter
true
in one sample t test if the means are not equal it means
that there’s different populations
two sample design is usually
an experiment
one value (or level) of the independent variable
treatment
a group to which other groups are compared
a control group
a group that receives treatment in an experiment and whose dependent variable scores are compared to a control group
experimental group
the measure of how different two things are
effect size
design in which scores from each group can be logically matched
paired-samples design
most common paired design
repeated measures
consists of two samples taken from the sample group
repeated measures
measured at two different times
repeated measures
design in which scores from each group can not be logically matched and research has no interest in matching the scores
independent samples design
a measure of the degree of difference
effect size
a measure of effect size for regression and correlation
R squared
a measure of effect size for the difference between two means
Cohen’s d
the resulting number is expressed in standard deviation units
cohen’s d
allows you to test for differences between more than two groups
anova
an inferential stat that compares means, compares variances and assessing interactions
analysis of variance
a NHST that allows one to test for differences between 2 or more groups with one independent variable
one way NOVA
using multiple t test would increase what?
type 1 error
theoretical distribution of F values
F distribution
what are F values?
they come from the F ratio
a ratio of variance
f value
the resulting number for the F ratio is the
F value or F stat
deviation of data and the mean
variance
the numerator for the F ratio consists of
the variance between two groups
for the F ratio, the denominator consists of variance within each group called
error term
when the F ratio is much larger than 1 (3 or greater) the null is
false
when the F ratio is near 1 the null hypothesis is
true
what should we reference to determine if the F ratio is large enough to determine if the null should be rejected
F distribution
sum of the squared deviations from the mean
sum of squares
the variance; a sum of squares divided by its degrees of freedom
mean square
the mean of all scores
grand mean
NHST of differences among means
F test
all the numbers in the experiment
X total
all the numbers from a treatment group
X treatment
number of treatment in the experiment (same as number of groups)
K
T/F: we dont know where the significance lies for F test (Where Group H is larger than group C)
true