Science in Medicine

Question 1

describe the categorial data types

Answer 1

nominal - category with no intrinsic value (sex)
ordinal - ordered categories (small/medium/large)

Question 2

describe the types of numerical data

Answer 2

interval - equal distance between each value, no true zero (degrees celcius)
ratio - same as interval but with a true zero (degrees kelvin)

Question 3

what tests are used for each population distribution?

Answer 3

normal - t-tests, ANOVA
skewed - non-parametric statistical tests (mann-whitney, kruskall-wallis)

Question 4

what are descriptive statistics?

Answer 4

summarises main features of dataset:
- central tendency
- distribution
- variability (measure = SD/SE)

Question 5

what is statistical significance?

Answer 5

statistical tests used to assess the probability that an observed difference means that there is a real difference between groups
statistically significant = unlikely to be by chance

Question 6

what are the 5 stages of hypothesis testing procedure?

Answer 6

1. create a null hypothesis
2. pick a statistical test
3. set desired significance level
4. run the statistical test
5. compare statistic with significance level

Question 7

what is a p-value?

Answer 7

probability of obtaining the observed difference (or larger one) given that no true difference exists

Question 8

what are type 1 and 2 hypothesis testing errors?

Answer 8

1 - wrongfully rejecting the null hypothesis
2 - wrongfully not rejecting the null hypothesis