Consider which data collection methods would be most appropriate for your proposed research and why. Restate your research  question and the methodological approach chosen for your proposed research  study. Then, identify and justify which measurement and data collection method best fits your research question and methodological approach. Please use the Learning Resources to support your post (i.e., cite and reference)

https://apastyle.apa.org/style-grammar-guidelines/…
DeRon Shepherd
Week 9
6301
Discussion 1
Research Question:
How does Autism affect different cultures in society?
Population and Sampling Techniques.
Autism is being diagnosed more frequently in Black and Hispanic children than in white kids in
the U.S., according to the Centers for Disease Control and Prevention. Among all U.S. 8-yearolds, 1 in 36 had autism in 2020, the CDC estimated. That’s up from 1 in 44 two years earlier.
But the rate rose faster for children of color than for white kids. The new estimates suggest that
about 3% of Black, Hispanic, and Asian or Pacific Islander children have an autism diagnosis,
compared with about 2% of white kids (Stobbe 2023). I want to focus on black male children
who are infants to the age of 10 years old. My sampling technique will be that of purposive
sampling. This type of sampling, also known as judgment sampling, involves the researcher
using their expertise to select the most helpful sample for research purposes.
Ethical and Cultural Concerns.
To guarantee that my study is considerate of participants’ cultural backgrounds and adheres to
ethical norms, I should demonstrate knowledge that guides practice with clients of various
cultures and be able to demonstrate skills in the provision of culturally informed services that
empower marginalized individuals and groups. Social workers must take action against
oppression, racism, discrimination, and inequities and acknowledge personal privilege (NASW,
2021). Black autistic people’s examination and diagnosis are significantly impacted by antiBlack racism. For Black children on the spectrum, misdiagnosis rates are relatively high. For
instance, compared to White autistic children, Black autistic children are misdiagnosed 2.6 times
more frequently, and this mistake is more likely to result in an adjustment problem or behavior
disorder (Mandell et al., 2007). Due to limited identification of those who require less support,
Black children are also more likely to be diagnosed when exhibiting clinical presentations of
ASD that call for more substantial support. This underscores the ways that anti-Black racism and
racial stereotyping affect accurate diagnosis for this population (Jarquin et al., 2011).
References:
Mandell, D.S., Ittenbach, R.F., Levy, S.E. et al. Disparities in Diagnoses Received Prior to a
Diagnosis of Autism Spectrum Disorder. J Autism Dev Disord 37, 1795–1802 (2007).
https://doi.org/10.1007/s10803-006-0314-8
Jarquin, Vanessa G. PhD; Wiggins, Lisa D. PhD; Schieve, Laura A. PhD; Van Naarden-Braun,
Kim PhD. Racial Disparities in Community Identification of Autism Spectrum Disorders Over
Time; Metropolitan Atlanta, Georgia, 2000–2006. Journal of Developmental & Behavioral
Pediatrics 32(3):p 179-187, April 2011. | DOI: 10.1097/DBP.0b013e31820b4260
National Association of Social Workers. (2021). Code of ethicsLinks to an external site.Links to
an external site..https://www.socialworkers.org/About/Ethics/Code-of-Ethics/Code-of-EthicsEnglish/Social-Workers-Ethical-Responsibilities-to-Clients
https://www.scribbr.com/methodology/sampling-methods/
Stobbe, Mike. (2023 March, 23). “Autism now more common among Black, Hispanic kids in the
US.” APSNews. https://apnews.com/article/how-common-is-autisme38179682e2759b0aff9c017bf7ebf61
Straiton D, Sridhar A. Short report: Call to action for autism clinicians in response to anti-Black
racism. Autism. 2022 May;26(4):988-994. doi: 10.1177/13623613211043643. Epub 2021 Sep
17. PMID: 34533064; PMCID: PMC9008552.
Yegidis, B. L., Weinbach, R. W., & Myers, L. L. (2018). Research methods for social workers
(8th ed.). Pearson.
•
•
Chapter 10: Measurement Concepts and Issues (pp. 223–245)
Chapter 11: Methods for Acquiring Research Data (pp. 246–276)
Measurement is also important in qualitative studies. Research- ers conducting qualitative studies
acknowledge, however, that their measurements may be biased by their own presence or other factors and will necessarily have a certain degree of subjectivity; they expect that. But that does not
absolve them of the responsibility to (1) measure even highly subjective variables, such as
impressions, perceptions, or the meaning of certain events to participants, as accurately as possible
and (2) measure more easily measured vari- ables, such as their participants’ demographic
characteristics or behaviors, accurately.
Most of us think we have a good understanding about measurement and how it works. After all, we
have used it almost all of our lives; we measure the height of children, our weight, room dimensions
and so on. But just to be sure we all agree on what it means, let’s review it in the context of
research. In research, measurement is the process of sorting and, when possible, quantifying
information in a consistent, systematic fashion. It entails collecting data in relation to certain
variables and assigning the appropriate value cate- gories or values to individual cases. In the social
work literature, there is disagreement as to how precisely one can measure certain variables. Some
writers have taken the posi- tion that, with enough effort, virtually all variables can be accurately
measured (Rubin & Babbie, 2014). They believe that most variables can be quantified, that is, a
number can be assigned to describe accurately the exact amount of the variable present within a
case. Other writers, especially those who espouse qualitative research methods, point out that
because of the nature of many of the variables that are of interest to us as social workers they
cannot be easily quantified using standardized measurement techniques. Our posi- tion is that, for
most (if not all) variables, a measurement strategy can be developed that will yield useful data for
researchers. However, we also acknowledge that many variables of interest do not readily lend
themselves to quantification.
Before any measurement can take place, two interrelated activities must take place. They are
referred to as conceptualization and operationalization.
Conceptualization
Conceptualization entails two major processes, (1) selecting and specifying what we believe to be
the most important or relevant variables to measure and (2) specifying as precisely as possible
exactly what we mean by those variables. In performing these tasks, we would rely on both
available knowledge (the literature) and our practice knowledge and experience. For example, the
literature may have suggested that abuse (or nonabuse) of medication by people who have a
terminal illness may be related to whether or not hospice care is pro- vided. It may have suggested
the hypothesis that terminally ill patients who receive hospice care will be less likely to abuse
medication than terminally ill patients who do not receive hospice care. Two variables that clearly
would need to be measured are (1) involvement (or noninvolvement) with hospice care, the
independent variable, and (2) the presence or absence of medication abuse (the dependent or
outcome variable). Specification of the independent variable (hospice
involvement/noninvolvement) would be no problem. But we would need to specify exactly what
we mean by medication abuse (the outcome or depen- dent variable). Would it be any use of
medication—accidental or intentional—other than exactly as prescribed? Should it include taking
less of a medicine than prescribed? What about the use of additional nonprescription medication?
All of the preceding questions (and more) would need to be addressed before data collection
begins, so that cases could be cor- rectly and consistently assigned to value categories for the
variable medication abuse. There are no right or wrong answers to these questions. We can answer
them any way that seems logical or, perhaps, just answer them in the same way that a previous
researcher addressed them in conceptualizing what he or she meant by the term “medication
abuse.” However we do it, the answers should be applied consistently with all research participants.
Sometimes, the answers are summarized at this point in what is referred to as a con- ceptual
definition. A conceptual definition tells the reader of a research report exactly how the variable was
defined in the current study. (Another researcher might apply still a different conceptual definition
to the same variable in his or her research.) A conceptual definition says, in effect, “When I use a
term in this research, this is what I mean by it.” For example, we might construct the following
conceptual definition:
Medication abuse is the intentional misuse of a prescribed, controlled narcotic sub- stance on at least three
occasions by a person who is terminally ill.
Generally, conceptual definitions are then followed by operational definitions. Opera- tional
definitions of variables are presented (along with definitions of other terms that may need to be
defined) at the end of a literature review or near the beginning of the methodology section of a
research report. In our example, it would also be necessary to define operationally such terms as
terminally ill and hospice care.
Operationalization
Operationalization is closely related to conceptualization and is the last step in prepara- tion for
measurement. It refers to specifying the actual measuring devices or methods that will be used to
measure key variables. Thus, it further clarifies their meaning.
Generally, there are many ways that a variable can be measured. For example, measurement of the
variable “medication abuse” might entail actual observations by nurses, social workers, caregivers,
or aides. It might also include social workers’ impres- sions based on mental alertness or other
symptoms of medication abuse, self-reports of patients, or something as specific as a measurement
of a certain chemical in the patient’s bloodstream at regular intervals.
We noted that operational definitions of a term may vary from study to study. Sim- ilarly, different
researchers may actually measure the same variable in different ways. Unless we are replicating the
research of others (in which case the same method of mea- suring the variable should be used), that
is fine. The only requirements for operation- alization of a given variable are that it should be
logical and justified (usually through the literature). Sometimes, researchers choose to include the
method of measurement (operationalization) of a variable in their operational definition of the
variable (Neu- man, 1997).
Levels (sometimes called scales) of measurement refer to the degree of precision with which a
variable is to be measured. We must decide at what level each variable is to be measured,
sometimes on the basis of the nature of the variable itself and more fre- quently on the method of
measurement used (the way it was operationalized). Deter- mining which of four possible levels of
measurement are present in the measurement of each variable is a critical step for deciding which
methods for any statistical analyses of data (see Chapter 13) would be appropriate. The four levels
represent a hierarchy of measurement precision from lowest to highest as described below. See Box
10.1 for a summary and examples of the four levels of measurement.
Nominal Level
The nominal level represents the most basic form of measurement. It involves the use of a
measurement scheme that simply sorts cases into exhaustive and mutually exclusive value
categories of a variable. In a nominal-level measurement, a case must fall into one (exhaus- tive)
and only one (mutually exclusive) category of a variable. The predictor variable in our previous
example (involvement/non-involvement in hospice care) would be an example of a variable that
would be considered nominal level. The different value categories of a variable ref lect only a
difference in kind, not a difference in the amount of the variable present within a given case. Even if
a number is used as a label for a value category to make data entry into a computer database easier,
the number has no quantitative significance. For example, in measuring the variable “gender
identity,” the researcher may assign the number 1 for individ- uals who identify as males and the
number 2 for individuals who identify as females, or the other way around. In either case, the
numbers do not ref lect a quantitative difference—just a qualitative one. They are merely labels,
substitutes for words.
In measuring many variables, nominal measurement is all that appears possible. Variables such as
gender identity, whether one voted in the last election, whether one owns a car or not, is an
undergraduate social work major, or one’s religious affiliation are nominal by their very nature. A
researcher might be interested in measuring the number of cars owned or degree of religiosity, but
those are different variables, ones that have the potential for more precise levels of measurement.
Level of Measurement
Nominal Ordinal
Interval Ratio
Brief Description
Categorical measurement; order of categories is not important
Categorical measurement; order of categories is important and cannot be arbitrarily changed
Numeric measurement; either does not have a possible value of 0 or it is not an absolute 0
Numeric measurement; does have an absolute zero value, meaning a value of 0 indicates there is none of the
variable being measured
Examples
Race, gender, marital status sexual orientation, ethnicity
Customer satisfaction (extremely satisfied, satisfied, dissatisfied, extremely dissatisfied)
Beck Depression Inventory Temperature Fahrenheit IQ Score
Number of children
Number of years of formal education Annual income
Type of Reliability
Reliability
Test-retest reliability Parallel forms reliability Split half reliability
Coefficient alpha Interobserver agreement
Question(s) Addressed
Does the measurement produce the same results under various conditions? Is the measurement consistent?
Does the measurement produce the same results when the test is administered two times to the same group
of participants?
Does the measurement produce the same results when participants complete two or more equivalent forms
of the instrument?
When the measurement is divided into two halves and completed by a group of participants, do their scores
on one half correlate with their scores on the other half?
Do all of the pairs of items correlate with one another? In other words, for each participant, are each of their
answers on the instrument consistent with their other answers?
When two observers record observations during the same period of time following the same instructions,
how closely do their recordings agree?
Tasks Required
There are many tasks that require special attention in research that employ secondary analysis of
data. They include the following:
•
•
•
Operationalizing the variables. Methods used to measure variables in the orig- inal data
collection should be identified, and if at all possible, the same opera- tional definitions
should be applied in the current study. Also, an estimate of the degree of reliability and
validity (Chapter 10) of the original measurement should be made.
Specifying the sampling plan used. The source of data (case record, personnel file,
monthly statistical report, and so on) and the strategy for selecting a sample of cases should
be specified and justified.
Developing a data collection instrument and coding scheme for data collection.
Typically, this is accomplished by using a data collection instrument that has been
developed specifically for the research. Data are drawn from the original documents and
•
•
recorded on the instrument. It is a good idea to have more than one person read and record
the same data from the original source. If this is done, an estimate of the interobserver
reliability of the data gathering method can be made (Chapter 10).
Analyzing the data. The appropriate level of measurement for each variable is determined
(on the basis of information about how it was originally collected) before statistical analysis
(Chapter 13) can be conducted.
Identifying the limitations of the study. Although most research reports typ- ically
include a section describing limitations of the study, the inclusion of a limitations section is
particularly important in secondary data analysis. Due to the nature of the data being
analyzed (they were measured and recorded for some purpose other than the current study
and usually by people other than the researcher), there is a greater-than-usual likelihood
that they will have limitations regarding their application to the current study.
Advantages
Secondary data analysis is appealing to researchers for a number of reasons. First, the
financial costs associated with it are often minimal compared with other data collection
methods. Second, it may require less time than other forms of data collection and anal- ysis.
The data are already collected and recorded; we need only to develop a sampling method
and an instrument for coding and recording them. If the data are already avail- able in a
form that is compatible with statistical analysis software, analysis can begin almost
immediately following the literature review and the development of focused questions
and/or hypotheses.
When ethically and legally feasible (under HIPAA, medical records are now gener- ally
inaccessible), secondary analyses are less intrusive than other methods that collect original
data from participants. Permissions may not need to be secured, as people are not
interviewed or observed in person by the researcher. However, when agency record data
are used, clients’ permission may be required because data were provided for use in
receiving assistance, not for research.
3. Develop a data collection instrument. In structured observation, instruments typically take the
form of behavioral checklists or category schemes. They are designed so that, when a behavior is
displayed, its characteristics can be easily recorded. ngagement
Behavior: Use empathy, reflection, and interpersonal skills to effectively engage diverse clients and
constituencies.
Critical Thinking Question: What steps would you need to take before you observed a child’s behavior in his classroom
from behind a one-way mirror?
instrument may also include demographic data on participants being observed, such as name, age,
gender, or other identifying characteristics. As is true for new data collection instruments, the
instrument should be pilot-tested before use. This is done to see how effective and accurate the
instrument is in describing the behavior.
4. Select an observer role. The observer role most appropriate to the behavior being studied is
selected. Logistical and ethical issues (as previously mentioned) are addressed.
5. Train observers. Structured observation requires the use of trained research per- sonnel. Unless
the researcher is conducting all observations personally (a potential source of bias), observers are
usually recruited or hired to observe. They need to be trained by the researcher in the method of
observation to be used and they need to gain experience with the use of the data collection
instrument. Observers need to know what to look for, when to observe it, how to know if the
behavior of interest is occurring, and how to record it using the instrument.
6. Conduct the observations. Observers observe participants’ behavior, recording the data as they
are obtained.
7. Verify the data. Data are checked for accuracy. One way to verify the data is to use two
observers to observe and record the participants’ behaviors. If the two observers generally agree
on the behaviors observed, the researcher can be reasonably confident that the data are accurate.