Flashcards in Midterm 2 Stuff Deck (41)
Loading flashcards...
1
True Positive
Condition positive and test positive
2
Type 2 Error
False Negative.
Test negative and Condition positive
-Reporting no effect when there is an effect
3
Type 1 Error
False Positive.
Test positive and Condition Negative
-Reporting an effect when there is no effect
4
True Negative
Condition Negative and Test Negative
5
How do we decrease false positives (Type 1 Error)?
Decrease Alpha level
6
How do we decrease false negatives (Type 2 Error)?
1) Increase Sample Size
2) Increase difference between means
3) Reduce variance
**Reducing false negatives increases the power of your study
7
Sensitivity
What percentage of the time the test correctly identifies the condition
True Positive/Condition Positive
8
Specificity
Percent of the time that the test correctly excluded the condition
True Negative/Condition Negative
9
Positive Predictive Value (PPV)
Proportion of positive tests that are true positives
"A positive test is a true positive X% of the time"
True Positive/Test Outcome Positive
10
Negative Predictive Value (NPV)
Proportion of negative tests that are true negatives
"A negative test is a true negative X% of the time"
True Negative/Test outcome negative
11
Accuracy
The percentage of cases that the test is correct
(True Negative + True positive)/All Tests
12
Prevalence
People With Condition/Everyone in sample
13
Can you predict a single outcome for a random event?
NO
14
Can you make any valid statements about a series of random events?
Yes
15
Central tendency Scores
-Ways of measuring the general area around which a set of numbers lie
-Mean
-Median
-Mode
16
Variability Scores
-Standard Deviation
-Range
-confidence Intervals
17
Who do you want to generalize to?
The theoretical population
18
What population can you get access to?
The study population
19
How can you get access to them?
The sampling frame
20
Who is in your study?
The sample
21
Standard Error of the Mean (SEM)
s/sqrt(n)
-smaller than standard deviation
-uncertainty in the estimate of the mean
22
Standard Deviation
Variability in Population
23
T-tests
Assumptions:
-Sample represents the population
-normal distribution
24
Analysis of Variance (ANOVA)
-Extension of t-test
-more than 2 groups
25
boufferont technique to adjust alpha
Alpha(EW)=Alpha divided by the number of comparisons
26
Repeated Measures Anova
Do two or more groups change differently over trials (over time)?
27
Correlation
-Relationship between two or more characteristics
-Correlation is a necessary but not sufficient condition for causation
28
Key Parameters for Correlation
r-correlation coefficient
r^2-coefficient of determination
p-probability that this relationship is not significant
b-slope
a-intercept
29
Best Straight Line Fit
Minimize sum of squared difference between data and curve fit line
30
Multiple Regression
Correlating more than one of predictor with a criterion
31
Power Analysis
-concerned with false negatives
-more subjects increases power
-more power is needed to detect small differences
-larger t=more power
32
Beneficience
Research participants should have a greater benefit than a risk
33
Autonomy
Freedom to withdraw or give consent at any time
34
Justice
Fair selection of research participants, whole population should be represented
35
Information needed in planning research
-alpha
-mean difference
-variance
-power
-sample size
36
Dichotomous variable
Ex) Sex
Two possible groups and you are either in one or the other
37
Cutoff Score
Continuous variable needs a cutoff to become a dichotomous parameter
38
Belmont Report
3 major guiding principles:
-Autonomy
-Beneficience
-Justice
39
Role of the IRB
-Meet and review all research proposals
-Make decisions regarding study
-Follow up on studies
-implement policies, procedures and documentation for review and follow up
40
Who sits on IRB
Faculty, clinicians, community members
41