No way to tell which caused the difference o Extraneous Variable (EV) – uncontrolled variables that can cause unintended changes between groups
o Confounded – when you don’t know whether the IV or the EV caused the difference o Book example: Reading Comprehension o Standard method of teaching used with 1 st graders and the New method is used with the 2 nd graders. We find a difference between the two groups. What is the EV in this example? Grade level A. Difference (A = standard method; B = new method) between two groups with no confounder operating B. Difference between two groups when a confounder is present and has moved the groups closer together C. Difference between two groups when a confounder is present and has moved the groups farther apart Nuisance Variables vs. Extraneous Variables o Example 1 o IV: list of words with no slow synonyms; List of words only slow synonyms o DV: Reaction time o NV: Age of all participants in both conditions is random mix of 12-78 yr olds o EV: age could be alternate explanation for the difference if one group is all senior citizens and other college students
o Example 2 o IV: words flash on screen count letters; words flash on screen put them in sentence o DV: recall a list of words o NV: The elevator goes off randomly when people ride I during both conditions. So sometimes people are effected by it and sometimes they aren’t o EV: Condition 1 is studied on the day of a big conference and the elevator goes off the whole day. Condition 2 is studied on another day when no one goes to the third floor conference Controlling Extraneous Variables o GOAL: produce groups that are equivalent prior to introduction of the IV, & eliminate EVs and reduce the effects of NVs as much as possible o Randomization : Ensures that each participant has an equal chance of being assigned to any group in an experiment o Example: Level of motivation o Drawback : NEVER fully aware of the variables that technique controls so you don’t know how effective its been o Elimination : EVs are completely removed from an experiment o Example: facial expression across cultures; only showed face so there is no bias of fat vs. skinny, clothes, etc. o Constancy – EV is reduced to a single condition that is experienced by all participants (no longer variable – only one level) o Situation: Juan and Nancy are interested in which teaching method is best for teaching stats. There are 2 classes and Juan will teach one using Method A and Nancy will teach the other using Method B. What is the problem? • Is the difference due to IV=Method or EV=Teacher How can we solve this using constancy? • Only use one teacher – only has one level or value o Controls for nuisance variables as well Testing conditions and participant conditions are same then shouldn’t have wide spread scores o Balancing o Achieves group equality by distributing EVs equally to all groups o Situation: (Juan and Nancy example – two teachers) Juan 25 Method 1; 25 Method 2 Nancy 25 method 1; 25 Method 2 o Counterbalancing o
You've reached the end of your free preview.
Want to read all 13 pages?
- Spring '11
- external validity, internal validity, participants