Research Designs Comparison: Strengths and Weaknesses
Aysha Siddiqui
Quantitative
Reasoning & Analysis (RSCH - 6200Y - 2)
Burger, J. (2009). Replicating Milgram: Would
people still obey today?
Burger (2008) used a controlled
experiment design for his experiment in which he replicated Stanley Milgram’s
(1974) experiment. The original experiment by Milgram was conducted to
understand the concept of obedience. In the replication of the experiment, many
steps and tests were added to the original one to address all the ethical
concerns.
The research design is a
‘one-shot case study’ in which the results are collected from a single group of
at a single point of time. This is a kind of Pre experimental design which is
mostly used when no other design is suitable for the study. These designs are
weakest for internal and external validity (Frankfort-Nachmias & Nachmias,
2008). The one shot case is also aimed
at understanding a phenomenon that produced change in the past.
The strength of this design is
the flexibility and ability of the researcher to conduct their study with a
unique design. This design gives an option to a researcher to conduct a study
though they were unable to fit their research design into the mould of other
traditional designs.
Marques, S., & Lima, M. L.
(2011). Living in grey areas: Industrial activity and psychological health. Journal of Environmental Psychology.
Marques & Lima (2011) clearly explain their choice of using
the Quasi-experimental design to understand the affect of living in industrial
areas on psychological health. Their hypothesis is that people living in
industrial area have lower psychological health. The second goal of the study
is to understand the relation between the ‘perception of living are as
industrial and psychological health’.
The weakness of Quasi-experimental design is that it allows random
selection but not random assignment (Frankfort-Nachmias &
Nachmias, 2008) and this was the weakness in this study as well.
The strength of a Quasi- experimental design is that they allow
researcher to use natural settings and in real life. This is the strength of
the study done by Marques & Lima (2011) as well when they are able to use a
sample of population living in ‘industrial areas’.
Recommend a quantitative design for your research plan.
The recommended quantitative
design for my research plan is the Quasi-experimental design. This design
allows random selection of sample but does not allow random assignment. Our research
will use the ‘planned variation design’ which measure ‘the casual affects of
systematically varied stimuli’ (Fraknfort-Machmias & Machmias, 2008). The
rationale for using this experimental design is:
- Most important reason to use this design
is because we are testing the variance in Intelligence Quotient (IQ) of
adolescents subject to their exposure to an activity.
Planned variation allows us to measure casual effects over a
period of time, and this is crucial in our research. We need to monitor the
time consumed by the kids in the activity for at least six months, and then see
it’s affects on their IQ.
- Past literature shows the use of this
design in studies similar to our research study. Frankfort-Nachmias &
Nachmias (2008) give the example of a policy- relevant study conducted to
check the affects of Head Start Planned Variation (HSPV) on the
development of academic skills among low income families.
- Our research will aim to distribute
important variables equally among the group of children chosen. Our first
groups will be kids, ages 7-11, who have spent in the past and shown
interest in playing any kind of ‘organized’ sports. The second group will
be the same age children who have exposure to playing video games (this
includes time spent on iPad, computer, Xbox, play station etc.).
- Planned variation design is suitable
mainly also because time is of essence here as our dependant variable is
IQ of the children. A review of past literature and theories shows that
children IQ changes over time (Flynn, 1998).
This may be due to some relevant factors that we
plan to keep as control variables.
For the designs that you did not choose, state why each one is not
appropriate for your research questions, hypotheses, and variables.
1. Classical Experiment Design
1. Classical Experiment Design
Frankfor-Nachmias and
Nachmias (2008) defined classical experiment design in which there two comparable
groups: an experimental and a control group. The experimental group is
equivalent to the control group, except the difference of exposure to the independent
variable. The experimental group is exposed to the variable and control group is
not. Pre test and post test measurement are taken of both groups and the data
is compared.
In our study we do
make a comparison, but there is not an experiment or control group. We do not
have two separate groups of sample based on a treatment or independent variable
exposure.
The comparison of two
groups is based upon exposure to the independent variable in the past and there
are two separate kinds of independent variables.
2. Cross sectional-
Not looking for opinions of people with or without control variables
Cross sectional
design is though the most popular design used by social scientists
(Frankfor-Nachmias & Nachmias, 2008) but it is not suitable for our research.
This design often uses survey research to collect information like past
experiences, backgrounds and attitudes. In most cross sectional studies
researchers are trying to find a relation between two variables.
In our research we
are interested in understanding the effects of independent variables, choice of
free time activity, on the dependant variable which is the IQ of the children.
Collection of data from the past will include test scores and time spent in
either sports or playing games. Our research is not aimed at understand the
opinions or background effects on present conditions.
Though we are aiming
to understand the relation between variables but there is an effect of one
variable on another.
4. Pre experimental
Design
According to Frankfort-Nachmias
& Nachmias Preexperimental designs are weak in internal and external
validity without allowing casual inferences. This design is usually used only
when no other design is suitable for a study. In our study, Quasi-experimental
design is suitable, so we do not need to consider pre experimental design.
References
Burger, J. (2009). Replicating
Milgram: Would people still obey today? American Psychologist , 64(1), 1-11.
Flynn, J. R. (1998). IQ gains over time: Toward finding the causes. The rising curve: Long-term gains in IQ
and related measures, 25-66.
Frankfort-Nachmias, C. &.
(2008). Research methods in the social sciences (7th ed.). New York: Worth.
Marques, S., & Lima, M. L. (2011). Living in grey areas:
Industrial activity and psychological health.Journal of Environmental Psychology, 31(4), 314-322. doi:10.1016/j.jenvp.2010.12.002
No comments:
Post a Comment