If you have any feedback for the author, E-mail me
REQUIRED TEXT: Experimental Methodology eighth edition by Larry B. Christensen, Allyn and Bacon, 2001 ISBN #0-205-30832-5 ------------------------------------------------------------ Chapter 8 Techniques to Control Extraneous Variables. Statistical control, e.g. ANCoVA Control during the experiment Randomization: extraneous variables known or unknown will not systematically bias the results. i.e. differential influence on experimental vs control groups. Random Selection: Representative sample = sample participants has the same characteristics as population. Each person has an equal chance of being selected and the selection of one does not affect the selection of another person. :. results can be applied to all of population. Equal an unattainable ideal Random assignment to IV groups of the randomly selected assures that the extraneous variables will not differentially effect levels of IV, i.e. be controlled. see page 200 for technique If cannot randomly select, then control extraneous variables by: Matching – increases sensitivity of experiment esp. when there are small differences since variables of the match are controlled. but (1)restricts size of the participant pool (2) results can be generalized only to match Alternatively, employ several levels of IV, additionally can examine interactions Yoked Controls – controls for temporal relationship between IV and DV. Precision Control – equates participants on several variables but difficult to measure and find articipants Frequency Distribution Control – participants selected so that each group has the save average scores on extraneous variable but never know is groups are really matched. Counterbalance – control for sequencing effects of order effect and carry-over effect 1) Intraparticipant – each participant experience one order of IV then its reverse, i.e. ABBA but, if effect is not linear, then ABBA BAAB 2) Intragroup – groups are counterbalanced rather than each individual, by incomplete counterbalancing, even though all possible sequences of IV are not tested, each level of IV is in each ordinal position & each level is preceded and followed by every other level Participant Controls: p 217 Double Blind Placebo Deception – provide participants an hypothesis unrelated to research hypothesis Present IV but measure DV at another time Ask participants their attitude toward experiment before, during and after tasks Experimenter's Bias: Recording errors => automation Equate attitudes and empathy Control expectancy error by Blind ------------------------------------------------------------ My office is in room 116C of HGH (826-4571) and I invite you to consult with me on any matter during my office hours or any time that is agreeable to you and me.Go back to the beginning
Copyright © 2002, Dr. John M. Morgan, All rights
reserved -
This page last edited Nov 25, 2002
If you have any feedback for the author, E-mail
me