# 3.1 intro to biostatisitics Flashcards

1
Q

The null hypothesis is either accepted or rejected based on what?

A

Statistical analysis

2
Q

What does the null hypothesis state?

A

Research perspective which states there will be no true difference between groups being compared

3
Q

What 3 primary levels for variables?

A

Nominal data, Ordinal data, and intervel/ratio data

4
Q

What are the 3 key attributes of data measurement?

A

Order/Magnitude. Consistency of scale/equal distances. Rational absolute zero.

5
Q

What is an ex of nominal?

A

Labeled variables without quantitative characteristics. ex. Lawyer, truck driver.

6
Q

what is an ex of ordinal?

A

Theres order and magnitude. ex. Pain scale

7
Q

What is an ex of interval/ratio?

A

ex living siblings (#) and personal age (in years)

8
Q

Define mode

A

most frequent

9
Q

Define median

A

middle number

10
Q

Define mean

A

average number

11
Q

Define variance

A

Average of the squared differences in each individual measurement value x and the groups X-

12
Q

Define Standard deviation

A

Square root of variance value.

13
Q

What are parametric tests?

A

Stat tests useful for normally distributed data

14
Q
A
15
Q

Is the curve Negative or Positive skewed.

A
16
Q

Is the curve Negative or Positive skewed?

A
17
Q

If the data is skewed what cant it be?

A

Interval

18
Q

What is kurtosis?

A

Tells how peaked a value is

19
Q

If theres more cluster in a graph, what type of Kurtosis is it?

A

Positive

20
Q

If theres less cluster in a graph, what type of Kurtosis is it?

A

Negative

21
Q

What are the 4 key questions to selecting the correct statistical test?

A
1. What data level is being recorded. 2. What type of comparison/assessment is desired. 3. How many groups are being compared? 4. Is the data independent or related (paired)?
22
Q

What are the two sub catagory questions to ask for What Data LEVEL is being recorded?

A
1. Does the data have order/Magnititude? 2. Does the data have an equal distance?
23
Q

What are the 4 catagory questions to ask for What type of Comparison Assessment is desired?

A
1. Correlation. 2. Regression. 3. Surival 4. Group comparison
24
Q

What does Correlation (r) do?

A

Provides a quantitative measure of the stength and direction of a relationship between variables.

25
Q

What is partial correlation?

A

A correlation that controls for confounding variables

26
Q

For Pearson Correlation, a P>0.05 means what?

A

Means there is no linear correlation, there may still be non linear correlations

27
Q

How can Pearson Correlation control for confounding?

A

By being ran as partial correlation

28
Q

What is a Kappa statistic?

A

a correlation test showing relationship of agreement between/consistency of decisions determinations

29
Q

Pertaining to Kappa statistic, what do +1, 0, -1 mean

A

+1 classify everyone exactly the same way. 0 theres no relationship. -1 classify everyone exactly opposite of each other

30
Q

What is regression?

A

measure of relationship between variables by allowing the prediction about the dependent, or outcome, variable DV knowing the value/category of independent variables IVs.

31
Q

What are survival tests?

A

Compares the proportion of events over time, or time to events, between groups. Ongoing.

32
Q

What curve do survival tests get represented by?

A

Kaplan-Meier curve

33
Q

For Nominal data, when would you use Pearson’s Chi-square test?

A

2 groups of independent data

34
Q

For Nominal data, when would you use Chi-square test of independence?

A

More or equal to 3 groups of independent Data

35
Q

For Nominal data, when would you use Fisher’s Exact test?

A

Groups with more or equal to 2 groups with EXPECTED cell count of less than 5

36
Q

When would you use Bonferroni test of inequality?

A

When we need to adjust the p value for # of comparisons being made. Very conservative

37
Q

What are 3 key words for RELATED data?

A

Pre vs Post. Before vs After. Baseline vs End.

38
Q

For Ordinal data and interval data, What is the Student-Newman-Keul test?

A

Compares all pairwise compasrisons possible. All groups must be equal in size

39
Q

For Ordinal Data and interval data, What is the Dunnett Test?

A

Compares all pairwise comparisons against a SINGLE CONTROL

40
Q

For Ordinal Data and Interval data, what is the Dunn Test?

A

Compares all pairwise comparisons possible. Useful when all groups are not equal size

41
Q

For Interval data, what is the ANCOVA?

A

Compares the means of all groups or RELATED data against a SINGLE DV

42
Q

For interval Data what are two other Post-Hoc Tests for groups 3+?

A

Tukey, Sheffee, and Bonferroni correction

43
Q

What is the Tukey test?

A

slightly more conservative

44
Q

What is the Sheffe test?

A

Less affected by violations in normality and homogeneity or variances. MOST CONSERVATIVE

45
Q

What does the Bonferroni correction do?

A

Adjusts the p value for # of comparisons being made. Very conservative

46
Q

What are 3 required assumptions of interval/ratio data?

A

Normally-distributed. Equal variances, and Randomly derived and independent.

47
Q

For Interval/ratio data what is Levene’s test?

A

used to asses for variance that they are equal

48
Q

What 3 things do we do when Interval/Ratio data is not normally distributed?

A

Transform data to a standardized value. Use non parametric tests. Always run Descriptive stats and graphs.

49
Q

What is Type 1 error?

A

we reject null, but shouldve accepted. Meaning you states there is a difference when there is not.

50
Q

What is a Type 2 error?

A

accept null but shouldve rejected. Meaning You said there isnt a difference when there actually was.

51
Q

How would one minimize type 2 error?

A

Power

52
Q

The larger the sample size the

A

greater ability of detecting a difference if there is one. More power

53
Q

What are the 3 rules for Sample size determination?

A
1. Minimum dif between groups deemed significant. 2. Expected variationg of measurement. 3. Type 1 and 2 Erroo rated and confidence Interval.
54
Q

How is the p value determined?

A

Probability of observing a test statistic value as extreme or more extreme than actually observed if groups were similar.

55
Q

What do statistical tests determine?

A

Possible error rate or chance in comparing dif between variables

56
Q

Is the p value selected before or after a study starts?

A

BEFORE

57
Q

what value should P be to be statistical significant?

A

0.05 or less.

58
Q

What are 4 interpretations of a pre set p value?

A

Prob of making a type 1 error if null is rejected. Probability of claiming a dif between groups when there isnt. Prob of obtaining group differences as great or greater if the groups were actually the same/equal. Prob of obtaining a test statistic as higher if the groups were actually the same/equal.

59
Q

How are Confidence Interval CI calculated?

A

Calculated at an a priori % of confidence that statistically the real dif or relationship resides.

60
Q

What two things is CI based on?

A

Variation in sample (V/SD). Sample size (N)

61
Q

What is the interpretation of a 95% CI point estimate?

A

95% confident that the true difference or relationship between teh groups is contained within the confidence interval range

62
Q

What is the interpretation of a 95% CI without a p value?

A

If CI crosses 1 for ratios (OR/RR/HR) or 0.0 for absolute differences = NOT significant.

63
Q
A