Statistics

Question 1

What is a contingency table?

Answer 1

A summary of primary data that can be used to generate a Chi-square test to compare the observed results with expected results.

Question 2

What is discontinuous data?

Answer 2

Involves whole numbers only. Strict numbers.

Question 3

What is continuous data?

Answer 3

Involves decimal points. Flexible numbers.

Question 4

What is derived data?

Answer 4

Data calculated from findings. Mean, median, sum.

Question 5

What does descriptive statistics refer to?

Answer 5

The distribution of data. Look at the centre and the spread. Skewed or normal distribution (equal mean, median, mode)

Question 6

Define accuracy.

Answer 6

The closeness of measurements to the true value. BIAS indicates deviation from the true value.

Question 7

Define precision.

Answer 7

The closeness of REPEATED measurements to each other. Lack of precision indicates VARIABILITY.

Question 8

What is a T-test for?

Answer 8

To compare 2 groups and generate the P-value to accept or reject the hypothesis. Compares the mean of the 2 groups. They are used in continuous quantitative data. (decimal point)

Question 9

What is a two-sided hypothesis?

Answer 9

Looks at 2 angles for analysis. There is a DIFFERENCE between the test results from METHOD A & METHOD B. The difference can be in either direction; method A or B.

Question 10

What is a one-sided hypothesis?

Answer 10

Looks at one definite result. More or less. Results from METHOD A are higher than METHOD B.

Question 11

What is null and alternate hypothesis?

Answer 11

Null always states no difference or is always on the negative side. The alternate says yes.

Question 12

When to reject the null hypothesis?

Answer 12

When the P-value is <0.05%. such as 0.049%. Because the P-value is the probability value that the null hypothesis is true. So the lower the percentage the more unlikely it is true. Hence, reject the null (H0) hypothesis.

Question 13

What is the coefficient of variance? CoV.

Answer 13

It is the data indicating the sample’s S.D. in proportion to the sample’s mean. (S.D. / mean) x 100. It is expressed in %.

Question 14

What is the ANOVA test?

Answer 14

The analysis of variance. It tests if there is a difference (variance) in 2 or more groups by testing the differences in means of the groups. same as the T-test but for more groups.

Question 15

What is the standard deviation? S.D.

Answer 15

The standardised measure of the spread of data sets from the mean. How close values are from the mean. It is the square root of variance. square root (E(x-mean)2 / n -1)

Question 16

What is involved in an experimental design?

Answer 16

Hypothesis, experimental plan, sampling units, type of data to collect (scale), data analysis, final revised plan.

Question 17

What is the scientific and statistical hypothesis?

Answer 17

The scientific hypothesis is what we EXPECT to find out based on PREVIOUS WORK, the statistical hypothesis are PREDICTIONS related to what will be measured.

Question 18

What are the 2 types of experiments?

Answer 18

Manipulative where one or more FACTORS ARE ALTERED and observational which INVESTIGATES LINKS between variables in natural conditions.

Question 19

What problems can occur with sample size?

Answer 19

If the sample size is TOO SMALL then the null hypothesis cannot be rejected and if it is TOO LARGE then the subjects may be unnecessarily exposed to the risk of harm.

Question 20

What are the 4 outcomes of hypotheses?

Answer 20

Correctly/incorrectly retain/reject the null hypothesis. Type 1 error (false positive, 5%) is when the correct null hypothesis is incorrectly rejected and type 2 error (false negative) is when the false null hypothesis is incorrectly retained.

Question 21

How is power calculated?

Answer 21

Power = 1- beta
beta = probability of incorrectly retaining a false null hypothesis

Question 22

What are the variables to calculate statistical power?

Answer 22

Effect size (strength of the relationship between 2 variables), sample size (power increases when N increases), probability of alpha and the P value. Alpha is the probability of rejecting a correct null hypothesis. P value is the probability that what we found is not due to chance (affected by sample size).

Question 23

What power analysis can be conducted during the planning phase?

Answer 23

A priori analysis calculates sample size and sensitivity analysis calculates effect size.

Question 24

What power analysis can be conducted AFTER the experiment is conducted?

Answer 24

Post-hoc analysis. Calculates observed power.