Statistics Flashcards
What does bias describe in a trial?
A situation where one outcome is systematically favoured
Variations in definitions and classifications of bias exist.
What is selection bias?
Error in assigning individuals to groups, leading to differences that may influence the outcome
Subtypes include sampling bias and volunteer bias.
What is volunteer bias?
When subjects who are at risk are more or less likely to participate in a study
Example: A study on Chlamydia prevalence among students.
What is non-responder bias?
When those who do not respond to a survey may differ significantly from those who do
Example: People who did not respond to a dietary habits survey likely have poorer diets.
What is recall bias?
Difference in the accuracy of recollections retrieved by study participants
Particularly problematic in case-control studies.
What is publication bias?
Failure to publish results from valid studies, often due to negative results
Important in meta-analyses.
What is work-up bias?
Occurs when clinicians are reluctant to order gold standard tests unless new tests are positive
Can distort study results, affecting specificity and sensitivity.
What is expectation bias?
Observers may subconsciously report data that favours expected outcomes
Only a problem in non-blinded trials.
What is the Hawthorne effect?
A group changes its behavior due to the knowledge of being studied
What is late-look bias?
Gathering information at an inappropriate time, such as studying a fatal disease years later
What is procedure bias?
Occurs when subjects in different groups receive different treatments
What is lead-time bias?
Occurs when a new test diagnoses a disease earlier without affecting the outcome
What are the four phases of clinical trials?
0, I, II, III, IV
Each phase has specific goals and participant numbers.
What is the goal of Phase I clinical trials?
Safety assessment, determining side effects in healthy volunteers
What is the goal of Phase II clinical trials?
Assess efficacy with a small number of patients affected by a disease
What is the goal of Phase III clinical trials?
Assess effectiveness, typically involving hundreds to thousands of participants
What is the goal of Phase IV clinical trials?
Post-marketing surveillance to monitor long-term effectiveness and side effects
What is a confidence interval?
A range of values that likely includes the true effect of an intervention
Specified probability called the confidence level.
How is the confidence interval calculated?
Lower limit = mean - (1.96 * SEM); Upper limit = mean + (1.96 * SEM)
SEM = SD / square root(n)
What does confounding refer to in statistics?
A variable that correlates with other variables, leading to spurious results
What is an example of confounding?
Age may confound a study on aspirin and colorectal cancer if groups are not matched
What is correlation?
A test for the association between variables
What does the correlation coefficient (r) indicate?
The degree of correlation between two variables
r can range from -1 to +1.
What is the difference between correlation and regression?
Correlation tests for association; regression predicts values of dependent variables
What is linear regression used for?
To predict how much one variable changes when another variable is changed
What are the phases of drug development?
Phase 1, Phase 2, Phase 3, Phase 4
Each phase has distinct goals and participant criteria.
What does the GRADE system assess?
The quality of evidence in systematic reviews and guidelines
What does a forest plot illustrate?
Results from different studies, typically in a meta-analysis
What does a funnel plot demonstrate?
The existence of publication bias in meta-analyses
What is a box-and-whisker plot?
Graphical representation of sample minimum, lower quartile, median, upper quartile, and sample maximum
What is a histogram?
Graphical display of continuous data categorized into a number of categories
What is a scatter plot?
Graphical representation using Cartesian coordinates to display relationships between variables
What is a box-and-whisker plot?
Graphical representation of the sample minimum, lower quartile, median, upper quartile, and sample maximum.
What is the purpose of a funnel plot?
Used to demonstrate the existence of publication bias in meta-analyses.
Define a histogram.
A graphical display of continuous data where the values have been categorised into a number of categories.
What does a forest plot provide?
A graphical representation of the strength of evidence of the constituent trials in meta-analyses.
How is a scatter plot represented?
Graphical representation using Cartesian coordinates to display values for two variables for a set of data.
What does a Kaplan-Meier survival plot show?
A plot of the Kaplan-Meier estimate of the survival function showing decreasing survival with time.
What is a hazard ratio?
Similar to relative risk but used when risk is not constant over time.
What is incidence in epidemiology?
The number of new cases per population in a given time period.
What is prevalence?
The total number of cases per population at a particular point in time.
Fill in the blank: The formula for prevalence is _______.
incidence * duration of condition.
How does intention to treat analysis function?
All patients randomly assigned to one of the treatments are analysed together, regardless of whether or not they completed or received that treatment.
What is the normal distribution also known as?
Gaussian distribution or ‘bell-shaped’ distribution.
What percentage of values lie within 2 standard deviations of the mean in a normal distribution?
95.4%.
Define standard deviation.
A measure of how much dispersion exists from the mean.
What does NNT stand for?
Numbers needed to treat.
How is absolute risk reduction calculated?
ARR = CER - EER or EER - CER.
What is the difference between odds and probability?
Odds are a ratio of the number of people who incur a particular outcome to those who do not, while probability is the fraction of times an event is expected to occur.
What does power refer to in a study?
The probability of rejecting the null hypothesis when it is false.
True or False: Publication bias refers to the tendency for only studies with negative results to be published.
False.
Define relative risk.
The ratio of risk in the experimental group to risk in the control group.
What is sensitivity in the context of screening tests?
TP / (TP + FN), the proportion of patients with the condition who have a positive test result.
What does specificity measure?
TN / (TN + FP), the proportion of patients without the condition who have a negative test result.
What is the formula for positive predictive value?
TP / (TP + FP).
Fill in the blank: Likelihood ratio for a positive test result is _______.
sensitivity / (1 - specificity).
What does reliability in statistics imply?
Consistency of a measure.
What is validity in the context of a test?
Whether a test accurately measures what it is supposed to measure.
What does the null hypothesis state?
That two treatments are equally effective.
What is the p value?
The probability of obtaining a result by chance as extreme as the one observed, assuming the null hypothesis is true.
True or False: A Type I error occurs when the null hypothesis is rejected when it is true.
True.
What is the probability of making a Type II error called?
Beta.
Fill in the blank: The power of a study is expressed as _______.
1 - beta.
What is the null hypothesis accepted when it is false?
Failing to spot a difference when one really exists, a false negative
This is referred to as a Type II error.
What is the probability of making a Type II error called?
Beta
What does the power of a study represent?
The probability of correctly rejecting the null hypothesis when it is false
This is also the probability of detecting a statistically significant difference.
How can the power of a study be increased?
By increasing the sample size
What is a Type I error also known as?
Alpha
What is the formula for power in hypothesis testing?
Power = 1 - the probability of a Type II error
What type of data requires parametric tests?
Data that is normally distributed
Name two examples of parametric tests.
- Student’s t-test
- Pearson’s product-moment coefficient
What is the Mann-Whitney U test used for?
To compare ordinal, interval, or ratio scales of unpaired data
What does the Wilcoxon signed-rank test compare?
Two sets of observations on a single sample
What is the chi-squared test used for?
To compare proportions or percentages
What do normal (Gaussian) distributions satisfy?
Mean = median = mode
In a positively skewed distribution, which is greater?
Mean > median > mode
In a negatively skewed distribution, which is less?
Mean < median < mode
What are the key features of a randomized controlled trial?
- Participants randomly allocated to intervention or control group
- May have practical or ethical limitations
What is a cohort study?
Observational and prospective study following selected groups based on exposure to a particular agent
What is the usual outcome measure in a cohort study?
Relative risk
What is the purpose of a case-control study?
To identify past exposure to a possible causal agent for a condition by matching cases with controls
What is a cross-sectional survey also known as?
Prevalence studies
What is the highest level of evidence according to levels of evidence?
Ia - evidence from meta-analysis of randomized controlled trials
What does Grade A recommendation indicate?
Based on evidence from at least one randomized controlled trial
What is meta-analysis considered in terms of evidence?
The ‘gold standard’
What is the primary goal of meta-analysis?
To increase power and improve estimates of treatment effect or association
What can affect meta-analyses?
Publication bias
What does variance measure?
The spread of scores away from the mean
How is variance calculated?
Variance = square of standard deviation
