Lectures 4-6 Flashcards
What do inferential statistics quantify?
difference between two sample means = determine between two possible explanations
what is the probability of obtaining your sample data and statistics from a population in which the null hypothesis is true (no difference)
(α) p-value
large p-value
small difference between the means, null hypothesis = high probability of occurring
small p-value
large difference between the means, null hypothesis = small probability of occurring (accept HA
when is the null hypothesis rejected?
at p-value = 0.05 (significance level)
what type of error occurs when reject H0 when it was true
Type I error
what type of error occurs when accept H0 when it was false
Type II error
what represents the probability of making a type II error
beta probability β | β ≠ 1 - α where a=0.05
Inverse relationship between type I error and type II error
as the chance of making a type I error is decreased, chances of making a type II error are increased = why significance level is at 0.05
goodness-of-fit test
1-nominal-scale variable where the frequency is compared to an a priori ratio
what does the goodness-of-fit test analyze
the frequencies of occurrence within each of the categories of the variable
a pre-determined, already-established distribution
a priori ratio
(goodness-of-fit test) H0
there is NO DIFFERENCE between observed and expected frequencies
(goodness-of-fit test) HA
data does not match a priori ration – there are differences between observed and expected
test statistic
finds p-value || calculated to determine the probability of the observed frequencies conforming to the distribution established by the expected frequencies
(goodness-of-fit test) degrees of freedom
k - 1 where k = # of categories
Roscoe & Byar’s rule (R&B rule)
must be met to determine if the sample size (n) is large enough
(goodness-of-fit test) and (contingency tables) testing for continuity
DF (k-1) must = 1
contingency tables
2-nominal-scale variables || analyze frequencies of occurrence within each of the categories of both (i.e.: male/female) variables
what do the contingency tables test for?
they DETERMINE if the frequency of occurrence of the categories of one variable is INDEPENDENT of the other
expected frequencies between contingency tables vs. goodness-of-fit tests
goodness-of-fit = a priori expected values || contingency tables = a posteriori expected values
what is term to describe when variables obtained are BASED on data (i.e.:via calculations)
a posteriori
(contingency tables) degrees of freedom
(r-1)(c-1)
multidimensional contingency tables
deal with 3 or more nominal-scale variables
