Science of Practice Flashcards
Crossover trial
Each participant is given both the control and experimental treatment at different stages in time and in a random order
Case-control study
Studies past features of those with and without disease to find potentially causal agents
Preferred type of study to determine a relationship
Cohort study
Monitors patients over time to determine natural history of a disorder and identify long-term consequences
Cross-sectional study
Used to determine prevalence
Examines a population in a single ‘snapshot’ period
Ecological study
May highlight an association between the incidence of a rare disease and potentially causal factor in different groups
Open to confounding
RCT
Used for investigating the efficacy of a new treatment
Not suitable for identifying causal or risk factors
Fisher’s exact test
Used for SMALL numbers of participants
Used to establish the significance of results between two independent, un-correlated variables
Repeated measures ANOVA
Analyses repeated scores over time
e.g. patients giving repeated measures of scoring for their pain whilst using analgesics
Chi-square test
Used for LARGER sample sizes
Compares the statistical significance of categorical outcome variables e.g. yes/no, positive/negative
Non-paired - treats the measurements as if they are independent
Parametric pearson’s correlation coefficient
Measures the correlation between two variables
More appropriate to use when one or both variables have a NORMAL DISTRIBUTION
Doesn’t specify the direction of the relationship
Spearman’s rank correlation coefficient
Measures the correlation between two variables
More appropriate to use when one or both variables are NOT NORMALLY DISTRIBUTED or the relationship between the variables is not linear
Kendall’s rank correlation coefficient
Measures the correlation between two variables
More appropriate if any of the variables are NOT NORMALLY DISTRIBUTED
Mann Whitney U test
A non-parametric test used to compare two groups without making the assumption that the data is normally distributed
Two sample t-test
Compares the means of two groups of normally distributed numerical data and continuous data
i.e. two sets of number scores recorded on the same sample at different times under different conditions
