Assignment
Analyze the second article you selected in Unit 1, based upon this weeks reading. SEE DETAILS BELOW.
Using information gained from this week's readings, analyze the second article you selected in Unit 1.
• Summarize the type of research used in the article, as defined by the text.
• Describe the sampling procedure that was used to recruit the participants, as defined by the text.
• Describe the advantages and disadvantages of the method of sampling used.
• Evaluate whether the sampling procedure was appropriate for the type of research, and whether it allowed the researcher to answer the research question or questions.
• Post the persistent link for the article in your response. Refer to the Persistent Links and DOIs guide, linked in Resources, to learn how to locate this information in the library databases.
• Cite all sources in APA style and provide an APA-formatted reference list at the end of your post.
Details for Readings:
Leedy and Ormrod text to complete the following:
• Read Chapter 1, "The Nature and Tools of Research," pages 1-7, stopping at the heading "Tools of Research." This reading will provide background information on what research is and what it is not.
• Read Chapter 2, "The Problem: The Heart of the Research Process," pages 27-51. This chapter provides information on research problems and sub-problems, and research questions and sub-questions. It also discusses how to use scientific research literature in order to identify a research problem.
Research Concepts
Scientific Method
You are likely familiar with the steps of the scientific method-it is the basis for all scientific inquiry, from an elementary school science fair project to a scholarly study of the impact of a mentoring program on former gang involved youth.
Research Problems
Research begins when a problem is identified. These problems are modified into research problems, which form the basis of scientific inquiries.
• Problem 1: Managers needing to deal with workers' low morale.
• Problem 2: Helping adults recover from childhood sexual abuse.
• Problem 3: Helping hurricane survivors' social needs without bankrupting agencies.
These problems are large-picture mission statements, which orient you to the kind of study you will undertake, but they are far too unfocused to provide a basis for a realistic research project. You will want to move from the problem in the general sense to a statement of the research problem. In other words, a problem is not the same thing as a research problem. A research problem is often referred to as a gap in the literature. This means that a research problem is something that is not known, something that has not yet been researched. Recall from Unit 1 that the first component of scientific merit is that research must advance the knowledge in the field. The way to do this is to contribute new knowledge. Additionally, the knowledge that would fill the gap in the literature (or address the research problem) must be meaningful and important to your field both as a science and as a profession. For example, in Problem 1, the workers' morale problem, the statement of the research problem might be something like this:
Research Problem 1: It is not known whether a management-by-objective approach to morale problems will improve or worsen the workers' morale compared with a human-potential development approach.
Research Questions
The statement of the research problem should lend itself to translation into a research question that asks precisely what the study must answer in order to solve the research problem. The research question is a conceptual question, indicating the exact scope of the study. For example, the research problem above would generate the following research question:
To what degree is workers' morale affected by a management-by-objective (MBO) management approach compared with a human-potential-development approach (HPD)?
Inspect the sample research question and you can see that answering it will require a number of steps, which can be stated as sub-questions.
• Sub-question 1: How is the level of workers' morale affected by an MBO approach?
• Sub-question 2: How is the level of workers' morale affected by HPD approach?
• Sub-question 3: Is there a difference in the level of workers' morale between those that are exposed to an MBO approach and those exposed to an HPD approach?
Qualitative Research Questions
There is nothing inherently different about a qualitative research question, except for the fact that variables and the relationship among variables do not easily apply. Qualitative analysis does not measure things, so variability-based definitions and operational definitions are not relevant to it. But phenomena or experiences (the stuff of most qualitative analysis) can be named conceptually, and the research question must ask about those conceptual constructs.
Because qualitative research does not inquire into relationship between variables, words like "effects," "impact," "influence," and the like are seldom used. In a study of the experiences of African-American middle-class professionals of being racially profiled by police officers, one might see a research question like this:
How do African-American middle-class professional men feel about, think about, and integrate into their self-image the experience of being stopped by a police officer for no reason other than race?
In this version, the concepts involved are clearly identified: "African American middle class professional men's experience," "racial profiling," and "feelings, thoughts, and integration into self-image." A more broadly worded research question might ask this:
What is the lived experience of African-American middle-class professional men who have experienced racial profiling?
Notice that quantitative and qualitative research questions are stated as a question, address the research problem, and are answerable using research methodology; therefore, not all questions are research questions. They have to meet these criteria.
Hypotheses
Qualitative studies do not employ hypotheses, which are formal predictions about the outcomes of statistical analyses.
There should be one hypothesis or hypothesis set for each research question in a quantitative study. What this really means is that many research questions break down into sub-questions, just as the research problem may break down into sub-problems, and there should be a hypothesis set (of a null hypothesis and an alternative hypothesis) for each sub-question. For instance, hypotheses that correspond to the sub-research questions above could be worded this way:
Research question: How is the level of workers' morale affected by a management-by-objective (MBO) management approach compared with a human-potential-development (HPD) approach?
• Sub-question 1: How is the level of workers' morale affected by an MBO approach?
o Null hypothesis 1: The level of workers' morale is not affected by an MBO approach.
o Alternative hypothesis 1: The level of workers' morale is affected by an MBO approach.
• Sub-question 2: Is the level of workers' morale affected by HPD approach?
o Null hypothesis 2: The level of workers' morale is not affected by an HPD approach.
o Alternative hypothesis 2: The level of workers' morale is affected by an HPD approach.
• Sub-question 3: What is the difference in the level of workers' morale between those that are exposed to an MBO approach and those exposed to an HPD approach?
o Null hypothesis 3: There is no difference in the level of workers' morale between those who are exposed to an MBO approach and those exposed to an HPD approach.
o Alternative hypothesis 3: There is a difference in the level of workers' morale between those who are exposed to an MBO approach and those exposed to an HPD approach.
The hypotheses for research questions can be articulated basically as a statement of the research question. A statement of the research question in which there is no relationship between the variable is the null hypothesis, and conversely, a statement of the research question in which there is a relationship between the variables is the alternative hypothesis. A hypothesis should be stated as a testable prediction (not a question).
Sampling
Populations and Samples in Quantitative Analysis
In quantitative studies, your research methodology and approach (be it experiment, quasi-experiment, or non-experimental correlation) will dictate how demanding your sample selection and assignment methods must be. The question is whether the sample sufficiently represents the population from which it is drawn. You, the researcher, must decide ahead of time how representative of the population your sample must be, and then you must design the study to account for that.
For instance, in a true experiment (in the lab), because wide external validity is quite important, random selection and assignment are the best tools to get high external validity (representativeness). Consequently, your design will demand truly random selection of your sample and random assignment to groups. In a quasi-experiment, you generally try to design for either random selection of the sample or random assignment of participants into the groups.
Some correlation studies do not demand quite that high level of random selection or assignment, but you must determine yourself how widely you are going to try to generalize your correlations to the population, and design your selection procedures accordingly. The more external validity you require, the more you need to design for a representative sample of the population.
Populations and Samples in Qualitative Analysis
Qualitative designs are not usually preoccupied with matters of external validity. Their purpose is different: to achieve rich and textured knowledge about people's experience of some phenomenon or issue. As a result, qualitative analysis focuses more narrowly and deeply on "information-rich" participants, rather than on a larger and broader sample representative of large populations. Nevertheless, the question of whether you want to generalize your findings to a larger population (and who that population might be) remains critical in designing qualitative studies. Any intent to generalize one's findings means that the sampling procedures must be designed as stringently as in quantitative studies.
Describing Sample Size
Sample sizes are typically smaller in qualitative studies, but that general statement does not justify any particular sample size. Phenomenological studies, for example, often have samples as small as seven to 10 participants. However, these can be very weak studies. At the same time, interviewing hundreds of participants is unrealistic. Qualitative researchers balance their practical needs against the need to have a wide enough spectrum of experience to make for meaningful findings.
Sampling Procedures
There are various kinds of sampling procedures, such as random, stratified, cluster, convenience, purposive, snowball, and others. Several steps are involved in the actual sampling procedure: identifying potential participants, contacting that pool, recruiting or inviting their participation, and organizing the sample in terms of how the inclusion and exclusion variables or criteria will be measured or identified and how the participants are assigned to groups (if that is to happen) after they were accepted into the sample.
Articles:
1. Office for Human Research Protections- The Belmont Report
2. Sampling for qualitative research by Martin N Marshall.
3. Purposeful Sampling in Qualitative Research Synthesis by Harsh Suri.
Text Book: Educational Research Quantitative, Qualitative, and Mixed Approaches by R. Burke Johnson and Larry Christensen, Fifth Edition.