DPT1, EBP FINAL > Correlation + Reliability > Flashcards

Correlation + Reliability Flashcards

(78 cards)

Study These Flashcards
1
Q

EBP PROCESS

A
  • NEED- should grow from a need for information provided by the patient
  • QUESTION- the need should be transferred into an answerable question
  • SEARCH- the best evidence must then be consulted (literature search)
  • APPRAISE- the evidence must be appraised
  • APPLY- if deemed to be valid + applicable, it is applied to the patient
  • EVALUATE- the performance must be monitored + evaluated both for individual patients + groups of similar patients
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

association

A

how measures relate, they have a relationship
* correlation has some degree of association (direction + magnitude)
* provides limited information on reliability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

does correlation have association

A

has some degree of association (direction + magnitude)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

does association provide information on reliability

A

provides LIMITED information on reliability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

agreement

A

directly proportional/consistent
* perfect correlation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

describe agreement correlation

A

perfect correlation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

correlation

A

determines whether (p-value) + to what degree (magnitude) a relationship exists between 2 or more quantifiable variables
* are variables related?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

correlation-
things to consider

A

-positive or negative associations- DIRECTION
-strong or weak associations- MAGNITUDE

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

direction

A

positive or negative associations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

magnitude

A

strong or weak associations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Pearson product moment correlation (Pearson Correlation) measures what

A

measures the strength of association between continuous variables (ROM, 0-100 scale, etc.)
-aka magnitude

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Pearson product moment correlation (Pearson Correlation) ranges from

A

-1.0 to 1.0

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Pearson product moment correlation (Pearson Correlation)- PARAMETERS

A

-population parameter, “p”
-sample parameter, “r”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Pearson product moment correlation (Pearson Correlation)-
which parameter are we usually focused on

A

sample parameter, “r”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

see table on page 2 of Emily study guide + INTERPRETATION

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

pearson correlation (r) of 0.923 indicates…

A

strong, positive correlation
-because it is close to 1.0

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

significance (2-tailed) of 0.000 indicates…
(if correlation is significant at the 0.01 level (2-tailed))

A

p-value is less than 0.01

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

example of what to write for interpretation of correlations table question

A

“there is a strong, positive correlation between post-concussion symptom catastrophizing + anxiety (r = 0.923)”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

example of what to write for probability interpretation/statistical significance using p-value of correlations table question

A

“there is less than 1% chance of no association between post-concussion symptom catatrophizing + anxiety (p less than 0.01)”
*remember less than alpha is statistically singificant

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

p value less than alpha =

A

statistically significant

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q
A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

significance of correlation coefficient

A

-observed “r” is one of infinite number of possible correlation
-statistical significance is observed, NOT due ot chance
* statistical significance DOES NOT mean a strong relationship because it provides no information about the magnitude of the relationship

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

does statistical significance mean a strong relationship?

A

NO
-because it provides no information about the magnitude of the relationship

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

coefficient of determination

A

r^2

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
r only tells us what
whether correlation is strong or weak
25
what is coefficient of determination (r^2)
proportion of the variance (fluctuation) of one variation that is predictable from the other variable
26
r^2 is expressed as
a percentage -so it can be interpreted to explain variance
27
example of r and r^2
-r = 0.923 -r^2 = 0.852 or 85% -"85% of the total variation in y can be explained by the relationship between x and y"
28
full interpretation of r and r^2 for exam
-there is a (strong/weak), (pos/neg) correlation between x and y (r = ____) -there is ____% chance of no association between x and y (p less than or greater than ____) -x explained ____% of the variation in y (r = ____, r^2 = ____)
29
factors affecting r
-distribution, outliers -assumption is "r" is only valid for linear relationships or normal distribution -correlation is very sensitive to outliers- can use a different tecnique or remove outliers to see change
30
does presence of correlation indicate a cause-effect relationship?
NO
31
reliability
provides an idea of how much error is associated with a given test or measure
32
what has reliability?
EVERYTHING- just a certain degree of it -excellent reliability vs poor reliability
33
many reliability (agreement) estimates are based on
measures of correlation (association) -while also accounting for measurement error
34
low reliability = less/more measurement error
more
35
high reliability = less/more measurement error
less
36
types of measurement error
-systematic errors -random errors
37
systematic errors
-predictable, consistent, constant -able to be identified + usually can be corrected
38
examples of systematic errors
-improper calibration of machines -consistent use of improper landmarks
39
random errors
-unpredictable, inconsistent, variable -not always able to be identified, can be minimized, but not always eliminated
40
examples of random errors
-level of training in individuals -fatigue of rate -environmental factors -patient
41
see image of graphs for systematic errors vs random errors on page 3 of Emily study guide
42
set up operational definitions before ____
clinical encounters -to minimize disagreement -this allows others to replicate + interpret- "can i read this paper + do the same thing?"
43
4 types of reliability
-test-retest reliability -internal consistency reliability -interrater reliability -intrarater reliability
44
test-retest reliability
doing the same thing twice -looking for consistency across time -period of time between varies but it is suggested to be completed within 48 hours
45
time suggested between test-retest reliability
retest within 48 hours
46
internal consistency reliability
-typically used when looking at reliability of questionnaries/surveys -important because there could be 2 different subscales
47
interrater reliability
consistency across mulitple testers
48
intrarater reliability
consistency wihtin 1 tester
49
quantifying reliability
effect size for reliability
50
reliability is expressed as which coefficient
"r"
51
reliability "r" ranges from
0 (no consistency) to 1 (perfect consistency)
52
general guidelines for reliability (will be provided on exam)
0-0.5 = poor reliability 0.5-0.75 = moderate reliability 0.75-1.0 = good reliability
53
continuous variables
numbers represent the meaning of measure -ex: ROM
54
what is used for continuous variables
-Pearson correlation -intraclass correlation coefficient (ICC)
55
Pearson correlation
to indicate association + strength of that relationship
56
Pearson correlation- interval scale
no true 0 -ex: temperature
57
Pearson correlation- ratio scale
there is a true 0 -ex: ROM, age
58
Pearson "r" is a measurement of
ASSOCIATION between ratings -NOT a true measurement if agreement
59
is Pearson "r" a true measurement of association
yes
60
is Pearson "r" a true measurement of agreement
NO
61
intraclass correlation coefficient (ICC)
preferred for continuous data because it reflects BOTH relationship + agreement
62
intraclass correlation coefficient (ICC) uses what model
ANOVA
63
by using ANOVA, intraclass correlation coefficient (ICC) separates error into
-subject variability -error variability -subject + error variability
63
subject variability
variability between individuals
64
error variability
variability within individuals for a different assessment
65
subject + error variability
total variability
66
intraclass correlation coefficient (ICC) compares what 2 things
compares individual error to total variability ratio -subject variability / (subject variability + errorv variability)
67
intraclass correlation coefficient (ICC)- ratio approaching 0 indicates...
NO AGREEMENT
68
intraclass correlation coefficient (ICC)- ratio approaching 1 suggests...
PERFECT AGREEMENT
69
categorical variables
numbers are often "coded" for meaning
70
categorical variables have ____ results
2 results -ex: yes or no, positive or negative
71
categorical variables- ways to establish reliability
-percent agreement -chance agreement -kappa coefficient
72
percent agreement
simplest way to establish reliability for categorical variables -is there agreement between 2 testers? or between tester + gold standard?
73
chance agreement
determined by expected proportion of agreement if the examiner's ratings were completely random
74
chance agreement for -2 choices -3 choices -4 choices
-2 choices = 50% chance agreement -3 choices = 33% chance agreement -4 choices = 25% chance agreement
75
kappa coefficient
BEST choice for reliability of categorical data because it considers chance + removes it
76
see calculation example for categorical variables on page 4 of Emily study guide
DPT1, EBP FINAL
flashcards
Decks in class (2)
# Cards
Brainscape helps you reach your goals faster, through stronger study habits.
© 2026 Bold Learning Solutions. Terms and Conditions