Which of the following is NOT a major assumption of the correlation coefficients

sample comes from a normal population

number of standard deviations above or below the mean

What does a non significant Levene test mean

did not violate the assumption of homogeneity of variance

Which of the following is NOT true of a systematic review

selection is not usually specified

Which of the following is true

higher positive likelihood test = more info for ruling in

running several separate univariate ANOVAs for each dependent variable increases

the probability of type 1 error

if the R2 does not change when variable X is added to the model in multiple regression than

variable X is not valuable in predicting

the coefficient of determination (r2) is looking at what

variables shared by two variables

studbud

Research Final

118 termsStudBud Administration26 plays

2nd half of the semester

Flashcards118 cardsFPractice quizPHost liveL

Questions 118

Answers

1
Which of the following is NOT a major assumption of the correlation coefficients
the relationship among variables is linear
both variables have enough variability
homoscedasticity
sample comes from a normal population
2
what is a z score
the amount of spread in the data
squared difference of scores from the mean
a single score that represents the whole data set
number of standard deviations above or below the mean
3
Which of the following is not a use of narrative research
provide practitioners with a broad base of information
to conduct a review when literature is sparse but diverse
produce reproducible results
introductory section of research papers
4
Which of the following likelihood ratios would indicate a small shift in probability
- LR 0.3
+ LR 5
+ LR 1
- LR 0.1
5
when doing a meta analysis which can you use to combine discrete data
weighted mean difference
t- test
Standardized mean difference
Petro method
6
when looking a repeated measures which type of T-test should be run
paired T-test
independent T-test
7
What does a non significant Levene test mean
violated the assumption of randomly selection
did not violate the assumption of random selection
violated the assumption of homogeneity of variance
did not violate the assumption of homogeneity of variance
8
Which of the following looks at Non parametric data
one way ANOVA
Repeated measure ANOVA
Friedman's ANOVA
9
What level of measurement must a non-parametric test be run on
interval
discrete
ordinal
ratio
10
With which type of review can you exclude articles you dont want to use (another language)
Systematic review
Narrative
11
TRUE or FALSE: Meta-Analysis has more power than individual studies
true
false
12
Which is not an assumption that needs to be met in order to run a parametric test
large effect size
level of measurement
random selection from a normally distributed population
homogeneity of variance
13
Which of the following is NOT true of a systematic review
selection is not usually specified
synthesis often includes meta analysis
often a focused clinical question
inferences are usually evidence based
14
Which correlation coefficient would you use when both variables are ordinal
spearman's rho
phi
pearson r
Cramer's v
15
TRUE or FALSE: a post hoc test is not necessary when running a MANOVA
false
true
16
TRUE or FALSE: incidence is new cases
false
true
17
T/F: the hypothesis for parametric tests needs to be more generalized than for nonparametric
true
false
18
Which of the following is NOT true about ICC
can be extended to accomodate more than 2 raters
depends on repeated measures analysis of variance (ANOVA)
looks at absolute and relative reliability
is designed for nominal data
19
which of the following is the relationship between scores taken by different testers
intrarater reliability
test-retest
interrater reliability
20
how many tails does a nondirectional independent t test have
2
1
21
if the the p-value set at the beginning was 0.5 and at the end was 0.3 than
your results are insignificant
your results are significant
22
which of the following methods is a symmetric distribution
f distribution
chi-squared distribution
t distribution
23
TRUE of FALSE: the term effective is used when looking at a diagnostic tool
true
false
24
which of the follow is the tally of the number of times each score is represented in a data set
variability
central tendency
frequency
25
Which of the following is not part of descriptive statistics
central tendancy
range
variability
frequency
26
Which is assessing the relationship between 2 variables with the effect of a 3rd eliminated
partial correlation
interclass correlation
multiple correlation
Canonical correlation
27
Is the pooled estimate of this study significant
yes
no
28
Independent t tests are often used to analyze what type of tests
pretest - postest
posttest only
nonequivalent control group
time series
29
TRUE or FALSE: Multiple regression can be accommodated for nominal data
false
true
30
TRUE or FALSE: EBM takes longer than systematic review
true
false
31
TRUE or FALSE: with ratios the numerator has to be a subset of the denominator
true
false
32
TRUE or FALSE: Correlation = Causation
false
true
33
Which of the following is true
higher positive likelihood test = more info for ruling in
higher positive likelihood test = more info for ruling out
higher negative likelihood test = more info for ruling in
higher negative likelihood test = more info for ruling out
34
What letter is most frequently used to represent the most common correlation coefficient
r
c
f
e
35
TRUE or FALSE the mean of the sampling distribution is assumed to be the mean of the population
false
true
36
running several separate univariate ANOVAs for each dependent variable increases
the probability of getting a non normal curve
the probability of type 2 error
the probability of type 1 error
the probability of needing to run non-parametric analysis
37
Which of the following effect sizes would lead to the most significant results
0.2
1.2
0.8
0.5
38
Which of the following do you not use for a parametric test
ANOVA
Mann-Whitney
T-test
mean
39
Which study is weighted the highest
Ha, 2010
Akin, 2013
Hemal, 2007
Julien, 2009
40
if the R² does not change when variable X is added to the model in multiple regression than
variable X is valuable in predicting
variable X is not valuable in predicting
41
Which would you run for linear regression on nominal data with >2 predictor variables
regression analysis
Log linear analysis
Chi-squared analysis
correlation analysis
42
when you create a ratio of incidence rate for one subgroup and the incidence rate for another
odds ratio
risk ratio
43
the coefficient of determination (r²) is looking at what
confidence intervals around the coefficent
variables shared by two variables
statistical significance of the coefficient
strength of coefficients
44
Which of the following is a "good" score on the pedro scale
10
7
5
3
45
which is when researchers identify groups/risk factors from the past then look at present state
nonconcurrent cohort study
concurrent cohort study
retrospective cohort study
46
in a mixed design two way ANOVA which component is the within subject factor
repeated
independent
47
what test is usually used to determine the statistical significance of the regression equation
t- test
F-test
Chi squared test
48
TRUE or FALSE: low statistical power is common in rehabilitation research
true
false
49
TRUE or FALSE: parametric tests are assumed to be more powerful than non-parametric tests
true
false
50
Which of the following is not a threat to statistical conclusion validity
violated assumptions
low power
low SEM
lack of clinical importance
51
Which of the following analyses can you do on a single subject design
t - test
chi squared
slop and variability analysis
Kruskal - Wallis test
52
Frequency can be reported in all of the following ways EXCEPT
histogram
stem and leaf
raw frequency with percentage
line graph
53
the number of ankle sprains in the month of september would be
point prevalence
period prevalence
54
TRUE or FALSE with correlation techniques variables must be defined as independent or dependent
false
true
55
Which do you use do appraise evidence of a systematic review for outcome measures
Pedro
MANOVA
COSMIN
QUADAS
56
TRUE or FALSE: Effectiveness is more standardized than Efficacy
false
true
57
TRUE or FALSE: a small SEE with linear regression is more precise
True
false
58
which is a nonparametric test used for the difference between 2 dependent samples
Friedman's ANOVA
Wilcoxon signed rank test
Mann-Whitnney test
Kruskal-Wallis test
59
T/F: Linear Regression is used to predict future characteristics from previously collected data
true
false
60
a negative correlation coefficient represents what type of relationship
inverse
direct
61
which is used to predict dichotomous outcomes from numerical independent variables
linear regression
factor analysis
multiple regression
logistic regression
62
Which method of multiple regression are putting all variables in and removing 1 at a time
Backward
Stepwise
Forward
63
Which correlation coefficient would you use with one continuous and one dichotomous variable
Pearson
Spearman
Point-biserial
Phi
64
Which of the following stats would you run if you want a MANOVA of two groups
hoteling T2
Hoteling-Lawley trace
Wilk's Lambda
Pillari-Bartlett trace
65
Which of the following is NOT true about Pearson Product Moment Correlation with Extension
cannot be used alone to assess reliability
looks at the magnitude of the difference between 2 measures
to extend reliability a paired t - test may be conducted
only measures relative reliability
66
a correlation score of what would indicate low inverse correlation
0.9
-0.8
0.4
-0.3
67
TRUE or FALSE: a smaller sample size will have a larger SEM (standard error of mean)
true
false
68
Which test would you do if you can use parametric on difference between 2 independent groups
independent T test
Chi Squared
Mann-Whittney
One way ANOVA
69
Which of the following is NOT used for reliability analysis
interclass correlation coefficient
Kappa Correlation
alpha correlation
pearson produce moment correlation
70
T/F: with many tests the overall probability of obtaining significant difference increases
true
false
71
TRUE or FALSE: Relationship analysis studies are usually non-experimental
false
true
72
TRUE or FALSE: the more participants you have the narrower the CI
false
true
73
If one of the variables has a restricted range how will the correlation coefficient be affected
appropriately high
artificially low
appropriately low
artificially high
74
Which of the following is NOT a principle of meta-analysis
Each study is weighted so some contribute more
summary statistics are collected for each trial
pt in one trial are directly compared with those in another
each trial is analyzed separately
75
Is the pooled estimate of this study significant
now
yes
76
TRUE or FALSE: variance means more than standard deviation
false
true
77
Which database has lots of data in context but is "dirty"
EMR
Administrative database
In house database
Abstracted MR
78
TRUE or FALSE: when doing a systematic review you must document how you searched the literature
true
false
79
which of the following do you NOT use for a non-parametric test
frequency
Kruskal-Wallis
standard deviation
Wilcoxon
80
which is the percent of individuals with a diagnosis correctly identified as positive
Sensitivity
Specificity
positive predictive value
negative predictive value
81
which of the following occurs with a smaller alpha
type 1 and type 2 error go up
type 1 and type 2 error go down
type 1 error goes down type 2 error goes up
type 1 error goes down, type 2 error goes up
82
For a disease with a short duration which of the following is true
prevalence is greater than incidence
incidence is greater than prevalence
83
Which is not one of the four major ways correlation coefficients are interpreted
confidence intervals around the coefficient
strength of coefficient
Variables shared by the two variables
future predictability of the relationship
84
which design is used to document the status of a group at a particular point in time
nonconcurrent cohort
cross - sectional
case - control
concurrent cohort
85
Based on the results of this study in relation to homogeneity what would you do next
perform fixed effect analysis
perform descriptive analysis
perform random effect analysis
perform subgroup analysis
86
TRUE or FALSE: with linear regression the mean of the residual should be 0
false
true
87
Which summary statistic do you used for continuous data
odds ratio
risk ratio
standardized mean difference
standardized error fo the mean
88
Which of the following is NOT true about a type 1 error
can reduce by increasing the sample size
probability of committing is alpha
can reduce by increasing stat power
there is no difference but you find one
89
TRUE of FALSE: alpha and beta in relation to error have an inverse relationship
false
true
90
With normally distributed scores what percent of scores are within 1 SD of the mean
75
82
68
50
91
TRUE or FALSE when doing a systematic review more than one researcher should select the studies
false
true
92
Overall is the research homogeneous
yes
no
93
Mean, median and mode are a type of
variability
frequency
central tendency
94
which of the following distributions is the distribution of squared z scores
F distribution
t distribution
Chi-squared distribution
95
Which of the following designs starts with the effect and looks for the cause
case - control
cross sectional
concurrent cohort
nonconcurrent cohort
96
TRUE or FALSE: alpha level is set after the commencement of a study
true
false
97
ANOVA uses which type of statistics
z statistics
f statistics
chi squared statistics
t statistics
98
TRUE or FALSE: Chi Squared analysis can only be used for nominal data
false
true
99
TRUE or FALSE: a good model consists of as many predictors as possible
true
false
100
Mode is typically used to describe which type of data
ordinal
interval
nominal
ratio
101
a special form of what test is used to determine the statistical significance of a coefficient
t- tests
Chi-Squared
Mann-Whittney
ANOVA
102
TRUE or FALSE: significant difference = clinically important difference
true
false
103
T/F:if an ANOVA F statistic analysis shows a significant difference a post-hoc test must be run
false
true
104
based on the I²value are the differences due to chance
yes
no
105
What is the rule of thumb for how many participants you need to approximate the population
30
1000
100
10
106
if two samples are similar what will their ranks sums look like with Mann-Whittney analysis
they will be different
they will be similar
107
TRUE or FALSE: if the use of an odds ratio is not valid then use a risk ratio
false
true
108
Based on the results of this study in association with homogeneity what should you do next
perform random effect analysis
perform subgroup analysis
perform fixed effect analysis
109
TRUE or FALSE: QUADAS is used for intervention trials
false
true
110
Which of the following is NOT a post-hoc test
Scheffe
bonferroni
Tukey
Kruskal-Wallis
111
a McNemar test looks at which type of data
ordinal
interval
nominal
ratio
112
A factor analysis is done for all the following reasons EXCEPT
theory testing
subject allocation
theory development
test development
113
When creating a variable code it should be no more than how many characters
8
50
20
15
114
Which of the following is attritioin bias
selective outcome reporting
blinding of outcome assessment
allocation concealment
incomplete outcome data
115
Which correlation coefficient would you use with nominal values with more than 2 categories
Kendall's tau
Kappa Coefficients
Craver's V
Phi
116
TRUE or FALSE: you might run a ANCOVA if the pretest scores are different between groups
true
false
117
what is the probability of a detected score occurred by chance, represents the true population
z-score
alpha level
p-value
effect size
118
if central tendency is skewed in what direction is it named
toward the hump
toward the tail

Research Final Flashcards | StudBud