STAT 100 Lecture Notes - Lecture 5: Dependent And Independent Variables, Clinical Trial, Randomized Response
STAT 100 – CHAPTER 5 & 6
Response variable – measures an outcome or result or a study.
Explanatory variable – variable that we think explains or causes changed to the
response variable.
• Ie. Independent or dependant variable.
Lurking variable – has an important effect on the relationship between variables
but it not an explanatory variable.
• 2 variables are confounded when their effects on a response variable
cannot be determined from one another.
Clinical trial – studies the effectiveness of medical treatments.
Double-blind experiment – experiment which subjects or experimenter recording
the symptoms do not know which treatment was received or administered.
Randomized comparative experiment – compares 2 or more treatments.
Random assignment -> Group 1 -> Treatment 1 -> compare treatments/results.
➔ Group 2 -> placebo -> compare treatments/results.
Logic of experimental design:
• Randomized comparative experiments produce groups of subjects that
should be similar on average in aspects of why treatments were applied.
• Comparative designs expose all groups to similar conditions, other than the
treatment received.
• Differences in response variables must be due to effects of treatment.
Principles of Experimental Designs:
1. Control the effects of lurking variables by ensuring all subjects affected
similarly by them. Then simply compare 2 or more treatments.
champagneclam160 and 37952 others unlocked
8
STAT 100 Full Course Notes
Verified Note
8 documents
Document Summary
Response variable measures an outcome or result or a study. Explanatory variable variable that we think explains or causes changed to the response variable: ie. Lurking variable has an important effect on the relationship between variables but it not an explanatory variable: 2 variables are confounded when their effects on a response variable cannot be determined from one another. Clinical trial studies the effectiveness of medical treatments. Double-blind experiment experiment which subjects or experimenter recording the symptoms do not know which treatment was received or administered. Randomized comparative experiment compares 2 or more treatments. Random assignment -> group 1 -> treatment 1 -> compare treatments/results. Group 2 -> placebo -> compare treatments/results. Principles of experimental designs: control the effects of lurking variables by ensuring all subjects affected similarly by them. Then simply compare 2 or more treatments: randomize to assign subjects to treatments so treatment groups are similar, use enough subjects in each group to reduce chance variation in results.