Study biases and Hypothesis testing Flashcards
What is selection bias?
Nonrandom assignment to participate in a study group.
Most commonly a sampling bias.
What kind of bias may be present in a study with many patients who are “lost to follow up”?
A selection bias - the disease may have early mortality, leading to a loss of certain participants.
What kind of bias may be present in a study with healy workers and volunteers?
Selection bias - Study populations are healthier than general population.
How can selection bias be reduced?
Randomization
Ensure the choice of the right comparison/reference group.
What is recall bias?
Awareness of disorder alters recall by subjects - common in retrospective studies.
Patients with disease recall exposure after learning about similar cases.
How can you reduce recall bias?
Shorten the time from exposure to follow-up.
What is measurement bias?
Information gathered in a way that distorts it.
Example: Hawthorne effect - groups who know they’re being studied behave differently.
What is procedure bias?
Subjects in different groups are not treated the same.
What is observer-expectancy bias?
Researcher’s belief in the efficacy of the treatment changes the outcome of that treatment.
Example: If observer expects treatment group to show signs of recovery, more likely to document positive outcomes.
How can measurement, procedure, and observer-expectancy bias be reduced?
Use of placebo control groups with blinding to reduce influence of participants and researchers on experimental procedures and interpretation of outcomes.
What is a confounding bias?
When a factor is related to both the exposure and outcome but not on the causal pathway.
Factor distorts or confuses effect of exposure on the outcome.
Example: Pulmonary disease is more common in coal miners than general population. However, coal miners are also more likely to smoke than the general population.
How can confounding bias be reduced?
Multiple/repeated studies.
Crossover studies (subjects act as own controls)
Matching (patients with similar characteristics in both treatment and control groups)
What is lead-time bias?
Early detection is confused with longer survival, seen with improved screening techniques.
Early detection makes it seem as though survival has increased, but the natural history of the disease is unchanged.
How can lead-time bias be reduced?
Measure “back-end” survival (adjust survival according to severity of disease at time of diagnosis).
What are the measures of central tendency?
Mean, median, and mode.
What is the definition of the mean, median, and mode?
Mean = (sum of values) / (total number of values)
Median = middle value of a list of data sorted from least to greatest
Mode: The most common value.
If there is an even number of values, the median will be the average of the middle two values.
What is standard deviation?
Stanrdard deviation = how much variability exists from the mean in a set of values.
What is the standard error of the mean?
How is it calculated?
SEM = an estimate of how much variability exists between the sample mean and the true population mean.
SEM - SD / (square root [n])
Where SD = standard devaition and n = sample size.
Thus, SEM decreases as n increases.
How is the shape of a normal distribution described?
In a normal distribution, how many of the values are one standard deviation away from the mean?
How about two?
Three?
Gaussian/bell shaped.
1SD from the mean: 68%
2SD: 95%
3SD: 99.7%
How is a bimodal distribution shaped, and what does it suggest?
Bimodal: Two peaks (two values with high frequency)
Suggests two different populations are being measured.
How is a positively skewed distribution shaped?
What will the relative size of the mean, median, and mode be?
Asymmetry with a longer tail on the right.
Typically Mean > median > mode
How is a negatively skewed distribution shaped?
What will the relative size of the mean, median, and mode be?
Asymmetry with a longer tail on the left.
Typically mean < median < mode
What are the null and alternative hypotheses?
Null hypothesis: There is no difference
Alternative hypothesis: There is some difference
What is type I error? What greek letter represents it?
α = probability of type I error
Stating that there is an effect or difference when none exists (null hypothesis incorrectly rejected in favor of alternative hypothesis).
AKA false positive error.
A p-value is judged against a preset α level of significance (usually < .05). If p < .05 there is less than 5% chance that the data will show something that is not really there.
α = you saw a difference that did not exist (e.g., convicting an innocent man).
What is type II error? What greek letter represents it?
β = probability of making a type II error
Stating that there is not an effect or difference when one exists. (null hypothesis is not rejected when it is in fact false).
β = you were blind to a difference that did exist (set free a guilty man).
What is statistical power? How is it calculated?
What variables affect power?
Power = (1 - β)
Power is the probability of rejecting the null hypothesis when it is false.
Increase power and decrease β by…
Increasing sample size
Increasing expected effect size
Increasing precision of measurement.
If you incresae sample size, you increase power. There is power in numbers.
What is a meta-analysis?
What are some of their limits?
Meta analysis: Pools data and integrates results from several similar studies to reach an overall conclusion. Increases statistical power.
Limited by quality of individual studies or bias in study selection.
What is a confidence interval?
How is it calculated?
Range of values in which a specified probability of the means of repeated samples would be expected to fall.
CI = range from [mean - Z(SEM)] to[mean + Z(SEM)].
The 95% CI (corresponding to p = .05) is often used.
For the 95% CI, Z = 1.96
For the 99% CI, Z = 2.58
When is the 95% CI for a mean difference between two variables not significant?
How about the 95% CI for odds ratio or relative risk?
Mean difference: When the CI includes 0, no significant difference, null (H0) not rejected
Odds ratio/relative risk: When CI includes 1, H0 not rejected.
When is there a significant difference between CIs between two groups?
When the CIs do nto overlap, significant difference.
When they do overlap, usually no significant difference exists.
What is a t-test, and what is it used for?
Checks differences between means of 2 groups.
Tea is meant for 2.
What is an ANOVA, and what is it used for?
Checks differences between means of 3 or more groups.
3 words: ANalysis Of VAriance.
What is Chi-square (χ2), and what is it used for?
Checks differences between 2 or more percentages or proportions of categorical outcomes (not mean values).
Pronounce Chi-tegorical.
What values is the pearson corrleation coefficient (r) always between?
How is it used?
r is always between -1 and +1. THe closer the absolute value of r is to 1, the stronger the linear correlation between the 2 variables.
Positive r value -> positive correlation
Negative r value -> negative correlation
Coefficient of determination = r2 (value that is usually reported).
What is primary, secondary, and tertiary disease prevention?
Primary: Prevent disease occurence (e.g., vaccination)
Secondary: Screening early for disease (e.g., pap smear)
Tertiary: Treatment to reduce disability from disease (e.g., chemotherapy)
PST: Prevent, Screen, Treat.
What does quaternary prevention refer to?
Identifying patients at risk for unnecessary treatment, protecting them from harm of new interventions.
What is medicare?
Who is it available for?
Federal program originating from amendment to the Social Security act.
Medicare is available to patients > 65 years old, < 65 with certain disabilities, and those with end-stage renal disease.
MedicarE is for Elderly.
What is medicaid?
Who is available for?
Federal program originating from amendment to the Social Security act.
A joint federal and state health assistance program for people with very low income.
MedicaiD is for Destitute
