Textbook Notes (369,067)
Canada (162,366)
Psychology (9,699)
PSYB01H3 (581)
Chapter 6

Chapter 6

5 Pages

Course Code
Connie Boudens

This preview shows pages 1 and half of page 2. Sign up to view the full 5 pages of the document.
Chapter 6: True Experiments I Basics of Experimentation  feature of an experiment is the degree of control a researcher has over it’s design  researchers directly control what levels of independent variables are manipulated  Experimental Control - exercised by holding constant as many extraneous variables as possible - relationship between independent and dependent variables cannot be confounded by a third variable provided this variable is held constant  Randomization - randomize unwanted effects that cannot be experimentally controlled either by holding them constant or by measurement - ex: memorizing a list of words → usually remember only the first few words (primacy effect) and the last words ( recency effect), not the middle → researcher would randomize the order of words so that each list has the same words but in different orders. - random assignment: used to assign participants to groups or to different levels of the independent variable ie. coin flip → get tails, go to group 1; ensure any extraneous influence is just as likely to influence one group as the other - psychology studies cannot use random assignment→ quasiexperiment: study where participants are not randomly assigned to an experimental and comparison group ie. comparing young and elderly participants (assignment by age, not random)  Independent Variable -directly and systematically manipulated by experimenter - referred to as factors -single-factor experiment: only one independent variable→ opposite of multifactorial designs - comparison group: gets a different treatment from the experimental group ie. in study to improve memory experimental group did the online brain training while comparison group just did computer-based training with no mental exercise - control group: doesn’t get anything → allows us to observe changes that happen in the absence of treatment  Dependent Variable - posttest: measurement of outcome of both groups after the experimental group has received treatment - pretest: measure the dependent variable prior to experimental intervention - using identical pretest and posttest is avoided  Control Variable - control var. can be a potential independent var. - control variables are used to show that these factors do not explain the result - other effects of other control variables that are not held constant in the study may be quantified/measured Research Design Alternative  between-subjects design: independent groups of participants receive the different levels of independent variable  within-subject design: all research participants receive all levels of the independent variable  Between-Subjects Design -different groups of participants must be assigned to each level of the independent variable - this design requires a large amount of people to enroll - often used in treatment outcome studies - good at making sure one treatment does not contaminate the other - concern is trying to make sure there are few differences among participants in the groups  Randomization and Matching - random assignment ensures groups are similar prior to manipulation of independent variable - neutralizes individual differences - pretest measure often used to evaluate whether random assignment was successful; check if groups are somewhat equivalent - randomized comparative change design: design with a pretest and posttest ie. memory study where a pretest was given before the treatment and a posttest afterwards to see the change -multiple posttests can be given to see how long the effect of a treatment is - dependent variable in some experiments cannot be measure in a pretest ie. studying effect of class size on student reaction to teachers → cannot measure their reaction until exposed to the treatment (different class sizes) - disadvantage of pretest: act of taking it can cause participants to change - random assignment removes bias but relies on chance which cannot guarantee perfect groups - the larger the group the less likely differences occur by chance - matching: used to equate groups by matching people from treatment and control groups (can be matched by age, gender etc) - unmatched characteristics may still influence results - when matching is used, the study becomes quasiexperimental - matched-pairs design: matching combined with random assignment; people are paired first and then randomly assigned to groups  Within-Subject Design - repeated measures: when the same participant can be tested repeatedly over time - efficient because performance can be observed across all experimental conditions - fewer research participants are needed - only used when experimenter is sure one treatment will not contaminate another - essential design for longitudinal studies - longitudinal research design: participants followed over time; able to collect data over decades ie. developmental psych studies - carryover effects: an error of this design where performance in one condition affects the performance in another - differences in treatment or between/among levels of independent variable may be masked by the order in which they are presented - carryover effects can also be caused by fatigued participants near the end of the study, and familiarity participants experience - randomization minimized carryover effects ( same idea as memorizing the list of words example)  Counterbalancing -technique used to control for the order of treatments - ensures each treatment or level of independent variable occurs in each time period of experiment - must calculate all possible treatment orders ie. for 5 treatments there are 120 different possible orders (5 x 4 x 3 x 2 x 1) - Latin-square design: an incomplete counterbalancing arrangement where each possible order of treatments occurs equally (pg.184 example of square) - there is still a possibility that the order of the treatments becomes confounded with the effects of the treatment - complete counterbalancing is the safest technique - experimental errors include participants have different answers to repeated testing  Selecting a Between- or Within-Subjects Design -if carryover effects are to be avoided then between-subject design is favoured - pharmacological investigations you between-subject designs because carryover would be fatal - for many psych studies individual differences are of interest ie. memory study would be within- su
More Less
Unlock Document

Only pages 1 and half of page 2 are available for preview. Some parts have been intentionally blurred.

Unlock Document
You're Reading a Preview

Unlock to view full version

Unlock Document

Log In


Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.