Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
29 Cards in this Set
- Front
- Back
Example of a clinical trial |
Does this medicine work? |
|
Ecample of nutrition and production studies |
Is the mean weight of animals on this diet higher than the mean weight of animals on this other diet |
|
Example of observational studies |
Is this breed predisposed to osteosarcoma |
|
Example of field studies |
Is there more disease in this region than other regions |
|
Examples of nominal level of measurement |
Mortality Incidence Prevelence |
|
Examples of ordinal level or measurement |
Scores of clinical severity Condition scores |
|
Examples of Interval measurement |
LWG SCC
|
|
Examples of ratio measurement |
SCC |
|
Examples of visual analogue scale of measurement |
Clinical severity |
|
How do we demonstrate differences in nominal data |
Differences between proportions |
|
Nominal data |
A set of data is said to be nominal if the values / observations belonging to it can be assigned a code in the form of a number where the numbers are simply labels. You can count but not order or measure nominal data. For example, in a data set males could be coded as 0, females as 1; marital status of an individual could be coded as Y if married, N if single. |
|
How do we demonstrate differences between ordinal data |
Difference between medians |
|
How do we demonstrate between interval and ratio and visual analogue scale |
Difference between means |
|
How do we demonstrate association |
Significance testing Errors of significance Statistical vs clinical significance One and two tailed tests Independant and related samples Parametric and non parametric techniques Hypothesis testing and estimation Sample size determination |
|
What is hypothesis testing |
Tests the null hypothesis -no difference Generates a p value |
|
What does a small p value mean |
Strong evidence of there being a difference |
|
What is a p value |
The probablility that the difference in sample parameters arose by chance |
|
What is a type 1 error |
Difference is shown in the results when there is not |
|
What is a type 2 error |
No difference shown when there is |
|
What is power? |
1-(type 2) The ability of a test to pick up a difference when there is |
|
Which type of error is of greater clinical importance? |
Type 2 |
|
What is the difference between clinical and statistical significance |
There may be a statistical significance in a result BUT it may not actually be clinically useful. i.e. a treatment lowers blood glucose from 20 - 19 and this is statistically significant but unlikely to use the drug as the animal is STILL diabetic |
|
What is the difference between one and two tailed tests? |
only one outcome - cant get worse if one tailed can get better OR worse if 2 tailed |
|
What do parametric tests focus on |
Parameters of normal distribution Mean and SD |
|
What is an independant and related sample |
Sample that is related may be due to this related change or external factors - e.g. discrimination by sex may mean a drug works/doesnt work BECAUSE of the sex difference not because of the drug itself |
|
What does Chi squared test measure? |
Estimates the probability of observed differences arising by chance, does not measure magnitude of association |
|
Why do you use t test |
Difference between means |
|
Why do you use chi squared test |
Difference between poropotions |
|
95% confidence is related to |
5% significance |