Stats related Flashcards
What is type 1 error?
occurs when we incorrectly reject a null hypothesis that is true (false positive) e.g. positive pregnancy test in a male
What is a type 2 error?
occurs when we fail to reject (accept) a null hypothesis that is false (false negative) e.g. negative pregnancy test in a pregnant woman
What is statistical significance (alpha)?
the probability of a type 1 error occurring. This can be pre-determined as the acceptable amount of Type 1 error or alpha. Most likely to be 0.05
What is study power?
Study power is the ability of a study to find a difference between the two arms.
It is predetermined to reduce type 2 error.
Most studies are powered to 0.8
What can researchers do to increase power?
Increasing power reduced type 2 error and this is achieved by:
1. Increasing sample size
2. increasing effect sizes (what difference in outcomes is seen to be representative of a difference)
3. increase significance levels (alpha)
What is a confounder?
a variable that influences both the dependent variable and independent variable, causing a spurious association. It can also be a source of bias
What does a 95% confidence interval convey?
95% confident that the true outcome effect lies between X and Y
What affects the confidence interval of a sample mean?
Standard deviation, sample size, and the type of sample data distribution
What studies are best suited for using odds ratios?
Case-control studies
How would you interpret a risk ratio of 2?
The exposure group has 2-times the risk of developing the outcome compared to the non-exposure group
How do you interpret a p-value of 0.05?
The chance of the observed difference occurring by random chance is 5%. A p-value less than 0.05 is often noted as being statistically significant.
When comparing baseline characteristics between two groups, what do you want your p-value to be?
As close to 1 as possible, to signify no significant difference in a baseline characteristic between the two groups
What is confounding by indication?
confounding caused by the indication of the intervention. For example, C-sections are 4.3 times more likely to result in maternal death compared to NVD. The confounding factor is the reason for a C-section.
True or False. Does standardization reduce confounding and other forms of bias?
True
What is a null hypothesis?
A null hypothesis assumes that any observed difference is a result of random chance and that no association exists
How are cohort studies different from case-control studies?
They are both examples of observational (non-interventional) studies
Cohort study you are comparing an exposure (most often a risk factor or disease) to a control (do not have the risk factor or disease)) and measuring their effect on an outcome. These are more likely to be prospective and measure effect as risk ratios
Case-control study you know what the outcome is (a disease for example) and you are trying to determine the effect of a risk factor. These are more likely to be retrospective and measure effect as odds ratios.
What is critical appraisal?
a systematic process used to identify the strengths and weaknesses of a research article to assess the usefulness and validity of its research findings
What are the six key components of critical appraisal?
(1) Assessing the appropriateness of the study design for the research question;
(2) Careful assessment of the key methodological features of this design;
(3) Suitability of the statistical methods;
(4) Accuracy of subsequent interpretation;
(5) identify potential conflicts of interest;
(6) Relevance to their surgical practice
What is the evidence hierarchy of medical literature?
Level 1:
- A: Systemic review of RCT
-B: High quality RCT
-C: RCT
Level 2:
- A: Systemic review of cohort studies
- B: Cohort study
- C: Outcome research
Level 3:
- A: Svstemic review of case control studies
- B: Case control study
Level 4: Case series
Level 5: Expert opinion
What is sensitivity?
The ability of a test to capture all individuals with a disease
What is specificity?
The ability of a test to exclude those who do not have a disease
What are the different types of randomization?
Simple, blocked, stratified, minimized
What is p-value?
The probability of the null hypothesis being true
What influences the p-value?
(1) Sample size (the larger a sample size, the more likely a difference will be detected);
(2) Magnitude of effect (a larger magnitude of effect will likely pick up a difference and therefore reduce the p-value;
(3) Spread of data (the bigger the standard deviation, the more spread, and the lower the p-value)
What is PRISMA?
Stands for Preferred Reporting Items for Systematic Reviews and Meta-Analysis. It is an evidence based minimum set of items for reporting in systematic reviews and meta-analyses via a 27-point check-list and flow chart
What are the six main components of the PRISMA checklist?
Interrogation of:
1) abstract
2) introduction
3) methods - eligibility criteria, databases used, search terms used, screening process for choosing studies, risk of bias assessment tool
4) results - use of PRISMA flow diagram including why studies were excluded, summary of the study characteristics + results, bias assessment outcome
4) discussion - interpretation of results, limitations of studies used, limitations of the review, implications of results
5) if a protocol was followed and where to find it
6) conflicts of interests
What are the benefits of PRISMA?
(1) Demonstrates the quality of the review.
(2) allows readers to assess strengths and weaknesses.
(3) permits replication of review methods.
What is incidence and prevalence?
Prevalence - the proportion of population with the disease in a time point
Incidence - the rate of new onset disease during a period of time
What are four types of bias?
Selection bias - sampling bias and response bias
Observation bias - failure to measure factors/outcomes accurately
Attrition bias - number of drop outs differs in separate arms, those who are left are not representative of the original study
Confounding factors - associated with exposure and outcome but independent.
What are mean, median, and mode?
Mean - sum of all values / number of values in the data set
Median - middle value of the data set
Mode - most common value of the data set
What is a meta-analysis?
A meta-analysis is the use of statistical methods to combine results of individual studies, usually RCTS.
It allows researchers to have a more precise estimation of treatment effect.
What is a systematic review?
Type of secondary research that involves collating all empirical evidence that fits pre-specified eligibility criteria in order to answer a specific research question.
It uses explicit, systematic methods to minimizing bias, thus providing more reliable findings from which conclusions can be drawn and decisions made
What are the advantages and disadvantages of meta-analysis?
Advantage
1. Greater statistical power to detect an effect than a single study
2. Combine results from several studies hence less likely to be influenced by local factors.
Disadvantages
1. pooling results from several studies may not predict the results of a single large study.
2. if the source is poorly designed, the meta analysis will produce unreliable results.
What is a Forest plot?
- It is used to demonstrate the results of meta analysis
- vertical line of no effect
- Diamond = point estimate of pooled
result - Horizontal line - 95% Cl; if CI crosses the vertical line of no effect, either sample size to too small or the result is not statically significant.
What is a Randomised control trial?
A research design that allows:
1. Comparison of 2 or more possible interventions
2. Participants are randomly divided to separate arms
Advantages:
- Minimise the risk of allocation bias
- Can be combined into systemic reviews/meta analysis to produce high level evidence
Disadvantages:
- Time and cost
- Difficult on long term FU
- Usually focus on a narrow area of patients condition
What are the different phases of RCT?
Phase 1: 50-100 healthy volunteers to evaluate drug safety and side effects
Phase 2: 100-300 patients with relevant illness to assess effectiveness and assess minimal dose
Phase 3: Large RCT thousands of volunteers to confirm drug safety + effectiveness when compared to a placebo or standard treatment
Phase 4. Postmarketing surveillance
