Trisha-Ann Matthew-Cadore
Education 703.22
Fall 2010
Assignment #1
Threats to Internal and External Validity
The action research will question the correlation between teaching methods and boys
achievement in the classroom. It will examine boys in a fourth grade common branch class from
a Title One school in Brooklyn, New York. The research will be in the quasi-experimental
design using the nonequivalent control group design model. Student surveys and tests in literacy
will be used to evaluate the effectiveness of the action research.
Internal Threats
History
There are many factors that can affect students performance and affect the data.
Weather conditions
Lateness, illness, absences
Students misbehavior
School interruptions: fire drills, assemblies and announcements
Insufficient nutrition: breakfast
Maturation
Students will be of mixed age and off course different maturity levels.
Testing
Students will not take a pre-test and a post-test for all tests. This will definitely affect the data.
Instrumentation
•The instruments will be valid because its tests that are used by the city. There are threats to
validity concerning the surveys because students might not be completely honest when
answering the surveys.
Selection
• The researcher is choosing the students who will participate in the survey.
Mortality
• Students may be discharge from the school for a number of reasons. This can threatened the
data.
External Threats
Participant Effects
• This can be a threat because the researcher does not know the students yet. As the researcher
gets to know them and their personalities, then students who bring certain types of behavior to
the action research will be discovered.
Ecological Validity
The researcher will be changing the students grouping. Boys seat together for a four week
period, which is different from what they are used to. This can be a threat.
Generalizable
The researcher does not think the action research can be replicated with the same results. This
can be a threat.
Pre-Test Treatment
Students may act differently due to being given a pre-test. This is a threat.
Selection-Treatment Interaction
The researcher is doing the selection. Permission is being granted from the principal and parents.
Placebo Effect- this is when the people being studied are basically sham into believing the
treatment they are given is the one to make them better. When in fact they are not being given
any type of treatment at all. Some patients feel better when they weren’t being treated at all.
Hawthorne Effect- Students may actually differently because they know they are being studied
and change their behaviors.
Compensatory Rivalry- This might be a threat because students might work extremely hard to
do well because they are motivated.
Novelty Effect- This can prove to be a threat to validity because there will technology used in
the classroom. Students may act differently because of the use of technology.
References
O’Connor-Petruso, S. (2010). Descriptive Statistics Threats to Validity [PowerPoint slides].
Retrieved from http://bbhosted.cuny.edu/webapps/portal/