Textbook Notes (368,317)
Canada (161,798)
Psychology (3,337)
PSYC 2360 (100)

Research Methods - Internal Validity.docx

6 Pages
Unlock Document

PSYC 2360
Naseem Al- Aidroos

Research Methods - Internal Validity Validity Threats 1. Construct validity 2. Statistical conclusion validity - whether or not you are drawing valid statistical conclusions 3. Internal validity - concluding validly that dependent variable is caused by independent variable 4. External validity - with what confidence can conclude that results generalize The Importance of Experimental Control  Occurs when extent that the experimenter is able to eliminate the effects on the dependent variable other than the effects of the independent variable o Rare to have something that is single-y caused in psychology  Greater the experimental control, the more confidently we can say that it is the independent variable rather than something else, that causes the dependent variable  Created through equivalence between groups  Extraneous Variables - has effect on dependent variable only and makes it seem that the independent variable is causing the dependent variable o Can still have internal validity o Constitutes random error  Confounding Variables - effects both the independent and the dependent variables o Problem of a third variable, can no longer conclude in valid fashion that independent variable causes the dependent variable Internal Validity  Independent variable causes dependent variable  what you can conclude if your experiment is internally valid  Internal validity is ensured only when there are no confounds  Confound  alternative explanation o Another variable is mixed up with the independent variable o Does not necessarily mean that the independent variable does not cause the dependent variable but always produces alternative explanations  And you don’t know which one is impacting dependent variable or if both were Extraneous Variables - People are Different  Problem of individual differences o Differences in disgust responses, response / reaction times o Probably not going to be able to control for all extraneous variables  Limited Population Designs o Select from limited and therefore homogeneous population (e.g. always using college students) o Typically used out of convenience but group is normally close in age, live in similar environments, have similar socioeconomic statuses, similar education background therefore there is reduced variability and limiting influence of extraneous variables  Before-After Research Designs o Often done in reaction time research and in clinical research  People act as their own controls o Any differences among participants will influence both the baseline memory measure and the memory measure that serves as the dependent variable o Dependent variable is measured at two different times and design increases power of experiment by controlling for variability among participants o Difference between this and repeated measures is that individual is only in one condition of experiment and first measurement of dependent variable is baseline o Problems?  Crates possibility of retesting effects  Fatigue, reactivity, develop strategies to do better on second test, increased likelihood they will guess research hypothesis StsentencesB using Recalllist B All participants Participants studyand recalllist randomlyassigned A to conditions Sinstructions no Recalllist B  Matched Group Designs o Participants are measured on the variable of interest before experiment and then are assigned to conditions on the basis of their scores on that variable o Reduces differences between the conditions on the matching variable therefore increases power of statistical test o Participants can also be matched through more than one variable o As long as matching variable actually correlates with the dependent measure, the use of matched group design will reduce random error and increase statistical power of the design o Not always necessary  Random assignment is sufficient to ensure no difference between groups  Matched groups is used when feel need to reduce variability within groups o Most effective when: measure is correlated with dependent measure, when there are expected to be large differences among participants on measure and when sample size is small Extraneous Variables - Staying Consistent  There are enough extraneous variables between participants before they come into the lab, why make more differences by treating participants differently?  Standardization of Conditions o All participants in all levels of independent variable are treated in exactly the same way with the exception of the manipulation itself o Hold constant every other possible variable that could possibly influence the dependent variable  How? o Experimental Script (protocol) - contact all participants in same manner  Contains all information about what experimenter says and does during experiment o Automation - if you can  Using automated devices (recorders, computers etc.) to run experiment  Allows for possibility of double blind experiment (experimenter and participant does not know what condition they are assigned to)  Does not allow participants to ask questions o Trai
More Less

Related notes for PSYC 2360

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.