12 Flashcards
What does it mean for a study to be experimental?
Assigns treatments randomly to individuals.
What does mean study to be observational?
Assignment of treatments is not made by researcher.
Reasons for experiments?
•eliminate bias
•Reduce sampling error
Placebo?
Eliminate Effect of treatment in which patient thinks the outcome is positive, and brings some bias by ensuring patient doesn’t know they have it.
Independent recovery?
Patients seek treatment when they feel the worst. This reduces bias.
Random assignment?
Averages out the effects of confounding variables. This can be done by randomly assigning treatments, which reduces bias
Blinding/anonymizing studies?
Preventing knowledge of experimenter of which treatment is given to whom. Bias is reduced.
Replication?
Carry out study on multiple independent objects to reduce sampling error.
Balance
Nearly equal sample sizes in each treatment, which reduces sampling error.
!!!Equation for standard error for t test, if noise increases in size, then?
The t test critical value goes along further in the distribution, making it easier to reject the null hypothesis
Blocking?
Extreme treatments?
Stronger treatments can increase the signal-to-noise ratio.
Matching?
Pair individuals in treatment group with control individuals with similar values for confounding variables. This reduces bias by limiting effects of these confounds, and reduces sampling error by grouping units into pairs.
Adjustment?
Use statistical methods to correct for effects of confounding variables.
Analysis of variance (ANOVA)?
•like a t test, but can compare more than two groups.
•Asks whether any of two or more means is different from any other
•In other words, is the variance among groups greater than zero?
