Download - Edad 510 final
Antonio Banks
EDAD 510
EDAD 510 Final Exam
1. Define the types of Internal and external validity threats in an
experiment.
By definition, threats to internal validity are problems that threaten our
ability to draw correct inference from sample data to other persons, settings, and
past/future situations. By structure, there are three main types of threats to
internal validity, each consisting of key factors.
The first types of threats are related to participants in the study & their
experience. Within this type, there are factors such as history, maturation,
regression, selection, mortality, and interaction with selection. History is
concerned with the time that passes between the beginning of the experiment
and the end, and the experiences that happen in between that can have an effect
of the eventual outcome. Maturation deals with the amount of growth in various
vectors that an individual may undergo during the process of the experiment.
Regression deals with individuals who start off in the experiment with a superb or
subpar pretest and over time have lower or higher score that regress towards the
mean. Selection deals with selecting individuals who may be more receptive to a
treatment, or more familiar with a treatment for experimental group. Mortality
deals with when an individual drops out of an experiment and drawing
conclusions for their scores is difficult.
The second type of threats to internal validity is related to treatments used
in the study. This includes diffusion of treatments, compensatory equalization,
compensatory rivalry, and resentful demoralization. Diffusion of treatments deals
with the communication across the control and the experimental groups occur.
Compensatory equalization refers to the need for both the experimental group
and the control group to both be granted benefits in some way, shape, or form
equally as opposed to one group (usually the experimental group) receiving the
benefits/treatment. Compensatory rivalry exists when the assignments are
disclosed to both the control and the experiment group; and the control group
feels as though they are undervalued in comparison to the control group. Lastly,
there is resentful demoralization, which occurs when members of a control group
may start to feel resentful n regards to their participation because they receive
what they perceive to be a less beneficial treatment than a control group
The third category of threats to internal validity is threats that occur during
an experiment and relate to the procedures. Types of these threats include:
testing and instrumentation. A threat to testing can occur when participants may
become familiar with the outcome measures and remember responses to for
later testing (Creswell, 2008). Instrumentation deals with threats that may occur if
the testing instrument is modified any time during the duration of the experiment.
By definition, threats to external validity are problems that threaten our
ability to draw the correct inferences fro the sample data to other persons,
settings, and past and future situations (Creswell, 2008). The types of threats to
external validity include interaction of selection and treatment, interaction of
setting and treatment, and interaction of history and treatment. Interaction of
selection and treatment pertains to the inability to generalize beyond the groups
in the experiment (such as other racial, social, geographical, age, gender or
personality groups) (Creswell, 2008). Interaction of setting and treatment deals
with the inability to generalize from the setting in which the experiment took
place. When this type of threat to external validity exists, the solution is often
times to analyze the effect of a treatment for each type of setting. Lastly, the
interaction of history and treatment deals with the researcher trying to generalize
findings to past and future situations (Creswell, 2008). Although there is no
method to accurately apply the research for a past situation, a solution for future
occurrences would be to attempt to replicate the study at a later date, as
opposed to generalizing the results for future situations.
2. Discuss measurement and characteristics of measurement, including
reliability and validity.
When developing or using an instrument for research it is important that you
understand the two key concepts of reliability and validity. Reliability pertains to
the fact that a testing instrument is stable and consistent when used across the
board (Creswell, 2008). The types of different reliability that exist include; test-
retest reliability, Alternative forms reliability, alternate forms and test-retest
reliability, inter-rater reliability, and internal consistency reliability. Test-retest
reliability is actively used when participants are asked to take a test, and then
they retake the test in constant intervals for reliability. In alternative forms
reliability, there are two separate instruments that are given to the participants
(some of the group getting test A, some getting test B), however they examine
the same types of variables; the researchers then correlate how the participants
did on both tests to test for reliability because the same concepts are being
tested. Alternate forms and test-retest reliability is a combination of the first two
test in which the first test is given, and then in a second retesting, the alternate
form is used for all participants. Inter-rater reliability is used when a group may
be making observations and are using a scale to answer questions about their
environment. Once the observation is completed, the group reconvenes, at which
time they share their scores with one another in order to make sure that the
scores are somewhat in line with one another, if not they have a chance to
discuss why they may have rated an aspect of the observation so high or low.
Internal consistency reliability is concerned with making sure someone would
complete items at the beginning of a survey in the same manner and method
they would complete it in the end. To check for such reliability, the Kuder-
Richardson split half test and the Spearman-Brown formula can be utilized.
The validity of measurement is also extremely important when conducting
a study. Validity ensures valid scores that could otherwise be skewed by things
such as: a poorly designed study, participant fatigue, stress, and
misunderstanding of questions on the instrument, inability to make useful
predictions from scores, poorly designed questions or measures of variables, and
information that has little use and application (Creswell, 2008). The types of
validity that exist within the confines of research include; content validity,
criterion-related validity, and construct validity. Content validity deals with how
well do the questions present possibilities of all the answers. If the question
doesn’t fully frame potential answers, there could be a loss of vital information in
translation that could have been vital to the study. Criterion-related validity
determines whether the scores from the instrument are good predictors of some
outcome they are expected to predict (Creswell, 2008). The two types of
criterion-related validity include predictive and concurrent. And lastly there is
construct validity, which is created by determining if scores are significant,
meaningful, useful, and have a purpose.
3. Define the purpose and use of correlation research and distinguish
explanatory and prediction designs.
Correlational research at its very essence serves a function of trying to
predict scores and explain relationships amongst variables (Creswell, 2008,
p.356). The function of correlational statistics is to describe and try to measure
the relationships that exist between any two given variables of the research. This
type of research is suitable when experiments cannot be performed (Schmidt,
2000), whether it is for ethical or practical reasoning. The key component for
correlational research with two variables is to determine the common variance
(Creswell, 2008). The key concept behind common variance is having the ability
to be able to make an educated guess about one of the variables by having
significant information pertaining to the other variable. In determining the
correlations that exist, there are two measures that must be considered. Direction
refers to whether or not the correlation is a negative or positive correlation
(Siegel,). In a positive correlation, it is seen that a positive variable has a positive
effect on the other variable; conversely a negative variable has a negative effect
on the other variable. Negative correlations can be seen when one variable has
an inverse effect on the other; a negative variable has a positive impact on the
other variable or a positive variable having a negative effect on the other
variable. The other measure that must be taken into account is strength of
relationship. The strength of the relation is quantified on a scale of 1.00, where
there is no relationship between the two variables, to -1.00, where the variables
have a perfect relationship, having a direct effect on one another.
Utilization of correlational research can be very useful when the main goal
of your research include the desire to observe the relationships between two or
more dependent variables. More importantly, if you want to determine if the
variables influence one another (Creswell, 2008) one of the unique features of
this type of research is the fact that it allows you to predict an outcome (Creswell,
2008). It is a very useful method when you know that running the correlational
statistical test will be of useful application in your work (Creswell, 2008). For this
type of research to ensue, there are two primary designs that are used; the
explanatory design and the prediction design. In explanatory research design,
the researcher’s primary goal is to explore in depth as to how the two or more
variables covariate to one another (Creswell, 2008). In using this design, there
has to be at least two variables to investigate correlation between, the
participants are usually analyzed as one single group, and the data is collected
for the research at one time. In prediction research design; the key motivating
factor is to determine which variable or variables are the predictor variables
(Creswell, 2008 ). A predictor variable is a variable that can be used to make a
highly educated guess regarding what is referred to as a criterion variable. For
example, the amount of time spent with a tutor may be a predicting variable in
determining an outcome such as course grade. Some of the key differences
between the two include the fact that in prediction research design, there are two
instances of data measuring/collection (explanatory has one). Another key
difference is that in explanatory research you usually make interpretations and
draw conclusions from the test results, whereas in explanatory research, future
performance is predicted based on the effectiveness of the predictor variables.
(Creswell, 2008).
Define ethnographic research, identify when to use it, and describe the
steps in an ethnography study.
An ethnographic research design is one that is centered around being
able to explicate, thoroughly document, and analyze the inner workings of a
particular culture. In order to do so, there has to be a significant time commitment
from the researcher(s) in order to achieve an accurate and detailed data set.
Things that must be accounted for in this type of research include being able to
describe, analyze, and interpret the culture’s shared patterns of behavior, beliefs,
and language (Creswell, 2008). In this type of research, the researcher or
research team must be meticulous in its approach to capture the needed data. In
order to gather this sensitive data, there is lots of interviewing, observing, and
document gathering (Creswell, 2008) that transpires. The work is a critical and
diversely utilized tool, described by Fetterman (1998) as “The art and science of
describing a group or culture, The description may be of a small tribal group in an
exotic land or a classroom in middle-class suburbia.”
The use of ethnographic research is useful in a variety of cases. In
summation, it can be used to explicate (with the permission of the group) a
culture’s foundation and day-to-day development. However it is stated by
Creswell (2008), that the research is extremely important when a better
understanding of that particular culture can have an impact on a far more vast
issue that may be taking place. For example, if you were to conduct a
ethnographic research around a patient of Alzheimer’s , including the patient
themselves, as well as the supporting cast (family, medical representatives, etc.),
you would gain a great knowledge base. You would begin to understand the
patterns that exists within their life cycle, their beliefs, as well as medical staff
and family, and you would be able to understand some of the language and
jargon that helps to exclusively deal with this issue. In turn, you will be able to
provide a accurate account of what may (not all), Alzheimer’s patients may
experience in their journey. There are three types of ethnographic studies that
can be performed; realist ethnographies, case studies, and critical
ethnographies. This research could provide a blueprint for further support for
research by other researchers. In conducting any of these three types of studies,
it is imperative that you pay close attention to the methods by which it should be
conducted. Five steps can ensure that a quality ethnographic study will be
conducted.
Initially, the researcher(s) must identify what the Intent of the research is
and determine the type of design that will be utilized. It is crucial that the
researcher understand explicitly what the intent of the research is and how they
believe it will have a viable impact on an issue. Once this is determined, it is key
to utilize the correct type of ethnographic study; the three all have differentiating
components that allow them to fit the particular culture you seek to study.
Secondly, one must discuss approval and access consideration. In doing this,
you are identifying the type of purposeful sampling you will be utilizing to answer
your research question. To do this, you must find an appropriate site as well as a
gatekeeper (Creswell, 2008). The gatekeeper’s function is to actively provide
access to your site and participants (Creswell, 2008). It is of extreme importance
that you do your best to be as little of a distraction to the site as possible and
provide a secure level of anonymity for all participants involved in your study. The
third step in this process is to make sure that you use the appropriate data
collection procedures in your project. In your data collection, you should be
intentional and exhaustive in your efforts. Rather it be the time spent building
rapport with the culture you’re examining, collecting various forms of data
through several outlets, or being a collaborator with the participants; you should
be very thorough, as it aids you greatly in the last two proceeding steps.
Sequentially, you must analyze and interpret the data within your design. You
should read through all of your data thoroughly (interviews, observations, field
notes, transcripts, etc.) , in order to help you develop a greater understanding of
the overall study. You must be able to identify emerging themes, as they can be
great indicators for your studies. Lastly, you must write and report your research,
which is consistent with the research you conducted. This report must be
objective in nature, and you must be able to keep your own personal opinions
and biases in the background of this report (Creswell, 2008). Additionally,
discussion at the end of the study should indicate how the research adds to the
knowledge base about the culture. (Creswell, 2008)