1|Page
NSG 522 MIDTERM EXAM NEWEST
ACTUAL EXAM WITH AND ANSWERS 100%
VERIFIED
statistical significance - correct answer_ the observed relationship b/w
samples = large enough to reflect a difference b/w pops. diff b/w pops
probably not d/t chance. if took more samples from the same pop,
you'd find that diff again.
too small sample can lead to... - correct answer_ T2 error!!
T1 error - correct answer_ p-value is stat'ly significant (<0.05), so you
reject H0 and accept H1. BUT there isn't really an r-ship. H0 was
incorrectly rejected! "you said there was a difference when there
wasn't one."
T2 error - correct answer_ p-value not stat'ly significant (>=0.05). don't
reject H0. BUT there really was a diff. "you said there wasn't a diff when
there really was."
,2|Page
alpha is.... - correct answer_ your cut point. level of significance. usually
set at 0.05.
p-value is..... - correct answer_ the RESULT of our test.
2-tailed test is... - correct answer_ almost always used. almost never
use 1-tailed.
The first step is always.... - correct answer_ ...dstats
Testing for the normality - correct answer_ Shapiro-Wilk
Bivariate test uses - correct answer_ -test for an r-ship
(assoc/correlations) b/w 2 vars (is the an r-ship?)
-make predictions using 1 predictor var & 1 outcome var (can 1 explain
the other?)
-compare groups using 1 ind var & 1 outcome var (is there a diff b/w
groups?)
, 3|Page
parametric tests - correct answer_ data is Perfect. (interval/ratio &
normal & no outliers)
non-parametric tests - correct answer_ nom/ord OR i/r data that's NOT
normal OR has outliers
steps in bivariate testing - correct answer_ 1) detm # of groups
2) detm if obs are ind (not from same subjs/related) OR dep (msmt of
same var in same person. ie. pretest/post-test)
NOMINAL data will be shown in... - correct answer_ cross-
tabulations/contingency tables. these examine rships b/w nominal &
ordinal vars w/ a small # of categories.
cross-tabulation/contingency table types - correct answer_ chi-square,
fisher exact test, mcnemar test. each test gives a p-value.
NSG 522 MIDTERM EXAM NEWEST
ACTUAL EXAM WITH AND ANSWERS 100%
VERIFIED
statistical significance - correct answer_ the observed relationship b/w
samples = large enough to reflect a difference b/w pops. diff b/w pops
probably not d/t chance. if took more samples from the same pop,
you'd find that diff again.
too small sample can lead to... - correct answer_ T2 error!!
T1 error - correct answer_ p-value is stat'ly significant (<0.05), so you
reject H0 and accept H1. BUT there isn't really an r-ship. H0 was
incorrectly rejected! "you said there was a difference when there
wasn't one."
T2 error - correct answer_ p-value not stat'ly significant (>=0.05). don't
reject H0. BUT there really was a diff. "you said there wasn't a diff when
there really was."
,2|Page
alpha is.... - correct answer_ your cut point. level of significance. usually
set at 0.05.
p-value is..... - correct answer_ the RESULT of our test.
2-tailed test is... - correct answer_ almost always used. almost never
use 1-tailed.
The first step is always.... - correct answer_ ...dstats
Testing for the normality - correct answer_ Shapiro-Wilk
Bivariate test uses - correct answer_ -test for an r-ship
(assoc/correlations) b/w 2 vars (is the an r-ship?)
-make predictions using 1 predictor var & 1 outcome var (can 1 explain
the other?)
-compare groups using 1 ind var & 1 outcome var (is there a diff b/w
groups?)
, 3|Page
parametric tests - correct answer_ data is Perfect. (interval/ratio &
normal & no outliers)
non-parametric tests - correct answer_ nom/ord OR i/r data that's NOT
normal OR has outliers
steps in bivariate testing - correct answer_ 1) detm # of groups
2) detm if obs are ind (not from same subjs/related) OR dep (msmt of
same var in same person. ie. pretest/post-test)
NOMINAL data will be shown in... - correct answer_ cross-
tabulations/contingency tables. these examine rships b/w nominal &
ordinal vars w/ a small # of categories.
cross-tabulation/contingency table types - correct answer_ chi-square,
fisher exact test, mcnemar test. each test gives a p-value.
