Choosing Statistical Test 1 Flashcards

1
Q

What must one ask when choosing a statistical test?

A

What is being tested? Are we comparing different samples for a point estimate (mean, median, frequency) - category 1
or are we demonstrating associations and relationships? - category 2

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

For category 1 questions, what must we ask?

A

What is the nature of the point estimate - mean or proportional?
How many groups are being compared?
How many observations of similar nature are made in each subject - paired or unpaired (independent) observations?
Can we assume parametric distribution - is the data not continuous?

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

Statistical tests for means

A

Parametric

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

Statistic test for means of one group

A

One sample test

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

Statistical test for means of two unpaired groups

A

Two sample test

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

Statistical test of means of two groups which are paired?

A

Paired t test

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

Statistic test for means with more than 2 groups which are unpaired

A

One way ANOVA

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

Statistical test for means of more than 2 groups which are paired

A

MANOVA

Repeated measures ANOVA

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

Statistical test for medians

A

Nonparametrical

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

Statistical test for medians of one group

A

Sign test

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

Statistical test for medians for two groups which are unpaired

A

Mann-Whitney U test

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

Statistical test for medians of two groups which are paired

A

Wilcoxon rank sum test

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

Statsistical test for medians of more than 2 groups which are unpaired

A

Kurska-Wallis test

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

Statistical test for medians of more than 2 groups which are paired

A

Friedman test

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

Statistical tests for proportions

A

Categorical

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

Statistical tests for proportions of 2 groups which are unpaired

A

Chi square test

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

Statistical test for proportions of two groups which are paired

A

McNemar test

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

Statistical test for proportions of more than 2 groups which are unpaired

A

Log-linear or logistic regression

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

What can parametric tests be used?

A

If at least one variable is quantitative and normally distributed

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

Which parametric tests are used when comparing groups?

A

T-tests

ANOVA

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

When are non-parametric tests used?

A

When either both dependent and independent variables are qualitative (nominal or ordinal)
or when variables are quantitative but not normally distributed

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

What happens in nonparametric tests?

A

Ranks are compared

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

What is the sign test?

A

Non-parametric test that compares median of a sample to median of the population

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

What is the non-parametric equivalent of the one sample t test?

A

Sign test

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

What is the Wilcoxon rank sum test used for?

A

Comparing two paired observations in nonparametric fashion

26
Q

What is Wilcoxon ranks um test equivalent to?

A

Paired t test

27
Q

What is Mann Whitney U test used for?

A

Independent observations in two groups

28
Q

What is Kruskal-Wallis

A

Non-parametric one-way ANOVA equivalent used for 3 or more groups

29
Q

Which statistical tests offer more robust analysis

A

Parametric

30
Q

Most commonly used method to transformation data from non-parametric to parametric?

A
Log transformation
Square root transformation
Square transformation
Reciprocal transformation
Logit transformation
31
Q

For what type of results does log transformation yield a normal distribution?

A

Right skew (positive)

32
Q

What is another name of normal distribution using log transformation?

A

Lognormal curve

33
Q

What can log transformation for right skew (positive) results do?

A

Yield normal distribution

Linearize a nonlinear correlation with upward shift

34
Q

What does square root transformation do?

A

Normalise and linearize

Stabilize variation

35
Q

What is square root transformation used for?

A

Normalising Poisson distributions

36
Q

What is reciprocal transformatino often used for?

A

Analysis of survival rates

37
Q

What is logit transformation often used for?

A

Distribution of proportions

38
Q

What does logit transformation do?

A

Linearizes a sigmoid curve
CI may be difficult to interpret after transformation
Geometric mean values may get distorted

39
Q

What is chi square test commonly used for?

A

Comparing frequency counts or proportions

40
Q

What type of test is Chi squared?

A

Non-parametric

41
Q

How does a Chi square test work?

A

Contingency table is constructed

Rows and columns are mutually exclusive

42
Q

What is a contingency table?

A

A table whose contents are in frequencies

43
Q

Classifications in contingency tests of chi square test

A

Observed frequencies = real outcomes observed at conclusion of study
Expected frequencies = outcomes predicted if null hypothesis was true

44
Q

What is chi-square?

A

Ratio of observed to expected frequency

45
Q

When is fisher exact test used in place of chi square?

A

If expected cell frequencies in more than 20% of cells falls less than 5

46
Q

When is Yates correction used in chi square?

A

If total sample is less than 100 of any cell value is less than 10

47
Q

When must chi square always be used?

A

Only for original frequency counts and not percentages

48
Q

What is the chi square test for paired data?

A

McNemars test

49
Q

What is the Mantel-Haenszel test?

A

A form of chi square test where the influence of binary categorical variables on one dependent variable is tested

50
Q

What is log-linear analysis?

A

Form of chi square test where log values of cell frequencies are employed

51
Q

When is log-linear analysis used?

A

When more than 2 groups or variables are studied

52
Q

Questions to ask if using statistical tests for category 2 data

A

How many dependent variables?
How many independent variables?
Are the variables categorical or continuous?

53
Q

Statistical tests for category 2 data

A

Testing association

54
Q

Test used if one dependent & independent variable, testing correlation in continuous, parametric data

A

Pearsons

55
Q

Test used if one dependent & independent variable testing correlation in non-parametric data

A

Spearmans

Kendall’s tau

56
Q

Test used if one independent & dependent variable testing regression in continuous data

A

Simple linear regression

57
Q

Test used if one independent & dependent variable testing regression using categorical data

A

Simple logistic regression

58
Q

Test used if one dependent variable and >1 independent variable testing regression in continuous data

A

Multiple linear regression

ANCOA

59
Q

Test used if one dependent variable and >1 independent variable testing regression in categorical data

A

Multiple logistic regression

60
Q

Test used if 2 or more dependent variables and >1 independent variable testing regression with categorical data

A

MANOVA

61
Q

Test used if two or more dependent variable and >1 independent variable testing regression using continuous data

A

MANCOVA

Multivariate multiple regression