click below
click below
Normal Size Small Size show me how
Experimental Design
Review of experimental design and vernacular
| Question | Answer |
|---|---|
| Used when the experiment has exactly two treatment conditions, subjects are paired off based on similarities - or act as their own "pair." Within each pair, each treatment is randomly assigned. Remember! Randomization occurs ONLY within the pairs. | matched pairs design |
| ________ occurs when control is not present in the experimental design. This happens when the researcher does not account for a variable that may have influenced the responses. The data collected in this experiment can't be used to draw conclusions. | confounding |
| _______________ bias occurs when subjects are unwilling or unable to participate in a survey. | nonresponse |
| When a subject responds to a treatment that is really neutral as if they have received an actual treatment, we call that the ___________ _____________. | placebo effect |
| When certain members of the population are not represented in the sample, we call that ___________________. | undercoverage |
| Used when the researcher divides the sample into subgroups based on a particular criteria, then assigns treatments randomly within the subgroups. | randomized block design |
| The researcher has NO CONTROL over any treatment or response but instead only records data. | observational study |
| When every member of the population is just as likely to be chosen as another. Every possible sample size n is equally likely to be chosen. | SRS (simple random sample) |
| In general ____ occurs when the results favor a certain outcome. | bias |
| _____ is when we ensure in our experimental design that multiple experimental units will receive each treatment. | replication |
| _________________ reduces bias and (hopefully) reduces the effect of lurking variables by spreading those variables out among the treatment groups. | randomization |