Pamela Minsch 7067 1 Lesson One Pamela Minsch William Howard Taft University Sept. 30, 2018 ED 511 Research Methods: Dr. Pamela Minsch 7067 2 1. Why is it important for educators to rely on research as a way of knowing or understanding? What are the limitations of such reliance? Educational research, according to the American Educational Research Association (2016), is “the scientific field of study that examines education and learning processes and the human attributes, interactions, organizations, and institutions that shape educational outcomes” (p. 5). According to Gay, Mills, and Airasian (2009), relying on research as a way of knowing is superior to using “experience, authority, inductive reasoning, and deductive reason to learn new things and draw new conclusions” (p. 5) unless these approaches are combined as part of the scientific method. As educators we are constantly seeking new and better ways to teach, to evaluate, and to connect with students. When we rely exclusively on these approaches to make these decisions, “the resulting knowledge is susceptible to error and may be of limited value to understanding the world beyond our immediate experience” (p. 5). For example, the authors point out that using inductive reasoning with a small sample population can lead to errors in our outcomes. Likewise, using deductive reasoning can lead to generalizations that are totally incorrect. ESSA, and NCLB before that, requires teachers to use research-based methods and materials to ensure the best outcomes for students. However, the authors describe four main factors that limit the use of research. First, some questions are just unanswerable. Second, the complexity of educational settings can be difficult to capture. Third, measuring instruments are not all perfect. Lastly, “the need to address participants’ needs in ethical and responsible ways” (p. 28) can limit the researcher’s ability to conduct a study. Pamela Minsch 7067 3 2. Describe a study you would be interested in conducting from a quantitative or qualitative approach. Briefly describe the problem, choose a research design, and identify the methods you would use to collect and analyze data. The Mat-Su Borough has a graduation rate of 81 percent (Mackintosh, 2018). Of the many factors affecting a student’s inability to graduate, Reschly (2010) suggests that difficulty with reading has been identified as one important factor. According to the author, early reading intervention and remediation is a key to student motivation and enthusiasm for school. Having taught at the middle school level for seventeen years, I wonder about those students whose reading scores are still low when they reach this level. If, for example, a student scores poorly on the state’s mandated, norm-referenced standardized achievement test, is that an accurate predictor of school drop out? And, if it is, should schools not be focusing on improving students’ reading abilities in middle school? To study this, one would need to design a correlational study, “a quantitative measure of the degree of correspondence” (Gay, Mills, Airasian, 2009, p. 9). The population for the study would be the graduating classes of two Wasilla area high schools for the last five years. One of the high schools is located in Wasilla, the other in Palmer; neither draws students from the same middle school (except for a few transfers). The method for selecting a sample would be Puposive sampling, “The process of selecting a sample that is believed to be representative of a given population.” It would be necessary to select a sample of four types of students from graduating classes: (1) graduates with high reading scores; (2) graduates with low reading scores; (3) drop outs with high reading scores; and (4) drop outs with low reading scores. The instrument needed to make a determination of correlation is the Measure of Academic Progress (MAP) test given to all 7th grade students since the passage of the No Child Left Behind Act. To determine if the MAP test can predict high school graduation, each of these four types of students needs to be Pamela Minsch 7067 4 studied. Because the study looks at data that is already collected and attempts to interpret the correlation of one variable, nonrandom sampling is the only choice for this applied research study. 3. The following are general educational research issues: (1) instructional leadership, (2) instructional strategies, (3) technology in the classroom, and (4) high stakes testing. Choose one of these issues and narrow it to a reasonable research topic and then a specific hypothesis. Defend the problem as researchable and evaluate the quality of your hypothesis. Of the four general educational research issues, technology in the classroom is one that is relevant to my situation. However, technology in the classroom is too broad a topic for research and, according to Gay, Mills, and Airasian (2009) “tends to result in a study that is general, difficult to carry out, and difficult to interpret” (p. 68). One aspect of technology in the classroom that is interesting to me is that of gamification. Gamification can be defined as the use of games as educational tools or it can be defined as the process of turning a learning environment into a game as Johnson, Adams Becker, Cummins, Estrada, Freeman, and Ludgate (2013) suggest: Gamification aims to incorporate elements of games, such as levels and badges (but also via quests and other strategies) into non-game activities. In gamified curricula, students can accumulate points or other rewards by accepting different challenges, and often have more freedom in choosing what kind of assignments they undertake to earn them (p. 5). A reasonable research topic for a quantitative study would be the extent to which the implementation of gamification in a given classroom increased the academic achievement of students. The specific hypothesis for this study would be: 7th grade social studies students who participate in a classroom structured as a multi-player game have higher end of unit assessment scores than 7th grade social studies students who do not participate in a gamified classroom setting. Pamela Minsch 7067 5 This study would extend the research on gamification as a motivational strategy in getting students engaged in learning. Many studies have been conducted on student motivation as a condition of engagement, but despite the growing interest in gamifying classrooms (Johnson et al., 2013) not a lot of research has been conducted on the benefits of gamifying curricula. The hypothesis is based on motivational theory, provides a reasonable explanation for the predicted outcome (Gay et al., 2009), and clearly states the expected relation between the variables. 4. You have been assigned the task of developing a research plan to investigate the effectiveness of using laptops in high school English classes. How would you organize a formal research plan? Identify the specific topics you would include in your plan and discuss your reasons for including each one. To investigate the effectiveness of using laptops in high school English classes, a formal research plan would need to include an introduction, a methods section, a data analysis section, a time schedule, and a budget section. Each of these sections is discussed in the following paragraphs. According to Gay et al. (2009) the introduction includes a statement of the topic, a review of related literature, and a statement of the hypothesis. For this study, the topic statement would be: The problem to be studied is the effect of using laptops on the quality of multiple paragraph essays in high school English classes. The topic is included to focus the research. The topic also guides the development of the hypothesis by identifying the variables; namely, the use of laptops and the quality of high school essays. For a literature review, ERIC and related educational databases would be searched for relevant prior studies. Including a literature review identifies what has already been written and helps “justify the significance of” (p. 80) the current study. Among other purposes, the literature review also ensures that the study will not duplicate one Pamela Minsch 7067 6 that has already been done. Finally, the hypothesis for this study would be: High school English students who use laptops to compose multi-paragraph essays will make significantly fewer convention errors than students who do not use laptops. The authors state that the hypothesis is a necessary component of the research plan because it determines all aspects of the study. Without the hypothesis this is no direction for the study. The method section includes the participants, the instrument(s), the materials or apparatus needed for the study, the design and the procedure. The participants and the larger population from which they are drawn must be identified. The instrument subsection discusses specific tools and how they will be used to measure the variables in your hypothesis. The materials subsection describes what, if any, materials or apparatus will be developed or used to conduct the study. The design subsection discusses “a general strategy or plan for conducting a research study” (p. 108). Finally, the procedure subsection describes how the data will be collected including the steps necessary for its collection. According to Kallett (2004), the methods section is the most important part of the research plan because “it provides the information by which the validity of a study is ultimately judged” (p. 1229). The data analysis section includes a description of how the researcher will analyze the data. Gay et al. (2009) suggest that the identification of “appropriate analysis techniques is extremely important” (p. 109). The importance of this is to ensure that the data can be analyzed prior to beginning the study. The time schedule section includes a list of the major activities and an expected completion date. It sets reasonable deadlines to keep the research moving forward and to give the plan’s reviewer an understanding of when the study will be concluded. Pamela Minsch 7067 7 Finally, the budget section is required for governmental or private agencies who fund research. 5. Choose two of the following probability sampling techniques: random sampling, cluster sampling, multi-stage cluster sampling, proportional stratified sampling, nonproportional stratified sampling, and systematic sampling. List the steps necessary to select a sample using each technique, and describe a situation in which you would use each technique. Two probability sampling techniques which can be used to ensure a reliable sample are stratified random sampling and cluster sampling. “Stratified random sampling divides a large population into homogenous subgroups called strata. Researchers then draw samples from each of the strata” (Allison, 2017, p. 1). On the other hand, cluster sampling is a method of sampling where “intact groups, not individuals, are randomly selected” (Gay et al., p. 129). The steps necessary to select a viable stratified sample similar to those for a random sample, according to the authors. The selection of the random samples, however, is done for each subgroup. According to Allison (2017), stratified random sampling begins “by dividing the population into homogeneous subgroups that do not overlap with one another” (p. 2). Gay et al. (2009) suggest that once you have subgroups in place there are two methods of sampling thee subgroups. These are proportional stratified sampling, where the size of each subgroup matches the percentage of that group in the overall population, and equal-sized sampling. For either, a random sample of the subgroups is selected using a table of random numbers. A reaearcher would use this method of sampling when the population to be sampled is scattered over a large area and the variables within the population, such as gender, age, years of experience, could affect the study. A second method of random sampling is cluster sampling. Cluster sampling selects groups rather than individuals are randomly selected for the study. The steps necessary to obtain Pamela Minsch 7067 8 a viable sample begin with the identification and definition of the population. Next, a determination of sample size must be made. The identification of “a logical cluster” (p. 130) and a list of all the clusters within the target population are the next steps. An estimate of the number of members per cluster and a determination of the number of clustered need to fulfill the sample size is made. Using a table of random numbers, the clusters re selected and “all population members in each selected cluster” (p. 130) are included. According to the authors, cluster sampling is used by researchers who cannot get a list of all members of a population, when the population is large, or when it is distributed over a large area. 6. A researcher is interested in describing the attitudes of teachers from rural, suburban, and urban schools across Louisiana. Recommend a random sampling technique to her, defend it as an appropriate way to meet her sampling needs, and describe the specific steps you would take to actually identify subjects. For an attitude study of teachers from rural, suburban, and urban schools across Louisiana, the researcher should use equal-sized groups in a stratified sample. One of the important variables in this study is the location of each of the schools as each represents a different type of community, community expectations, financial considerations, student populations, and other factors that makes each type of teaching situation unique. Because the researcher wants input from all of Louisiana, a large geographical area, and from three different types of teachers, the only method of that ensures that each group will be adequately represented is stratified sampling. The reason the researcher should use equal-sized subgroups is to obtain a large enough sample from each subgroup to make the study valid. Most likely there are many more urban teachers than rural teachers, so a proportional stratified sampling technique would mean that rural teachers were not adequately represented in the study. Once the researcher has decided to use stratified sampling, and already having identified his population, ie. all teachers from Louisiana, she must decide on a sample size. According to Pamela Minsch 7067 9 the authors, sample size is dependent upon the type of research and the size of the population, but if there are upwards of 5,000 teachers a sample size of 400 to 500 teachers would suffice. Next, deciding on how to organized the subgroups is necessary: for this study, the first level is rural, suburban, and urban. The second subgrouping should be elementary, middle, and high school levels (of course it could be private and public or some other grouping of interest to the researcher). Next, the researcher must identify each of the teachers as belonging to a specific subgroup. From these subgroups, she would randomly select his sample. 7. As the superintendent of a local school district, you are responsible for numerous accountability issues, one of which is the standardized testing program. You must decide on a test to be taken by all fifth graders in your district, but you are debating the merits of using a norm-referenced or criterion-referenced test. Discuss the advantages and disadvantages of each in this situation. There are two types of standardized achievement tests, norm-referenced and criterionreferenced. Norm-referenced tests compare the students who are tested to the performance of others who have taken the same test and provide a percentile rank. Criterion-referenced tests compare the students tested to a standard, giving specific information about specific learning objectives. Norm-referenced tests can tell a superintendent how the students in his district compare to others around the state, nation, or world, while criterion-referenced tests demonstrate how much growth a student has made since the last test was administered. There are advantages to both of these types of standardized tests. Schools and school districts need to ensure that the curriculum their teachers teach is educating students to the level necessary to be successful in college or a career. Norm-referenced tests give stakeholders a snap shot of how well a particular school or district is doing compared to others around the nation. Criterion-referenced tests tell the superintendent if the students in his district are moving along the continuum toward mastery of content. Pamela Minsch 7067 10 There are disadvantages of both of these tests as well. Norm-referenced tests are much more likely to be culturally biased as the one-size-fits-all basis of the test does not take into account the regional differences among students of differing locals. Students in Texas and California have a very different cultural background from those in Alaska’s rural villages. Even students in the same location can have cultural differences that make a norm-referenced test unfair. Criterion-referenced tests can have the same biases, depending upon where the test is developed. The biggest drawback for criterion-referenced tests is that their outcomes cannot be compared from location to location if they are not standardized like an AP assessment. Gay et al. suggests that another drawback is that criterion-referenced tests do not grade using the same bell curve as norm-referenced tests. 8. Explain why reliability is a necessary but not sufficient characteristic of a test. (p. 158 – 163) Also see summary pg. 169 9. You are conducting a survey study of teachers' views of the leadership style of their principals. You will sample teachers in 50 schools. You are aware that several of the principals are unpopular with their staffs. Describe how you will draw a sample of 1000 teachers and ensure that you receive an acceptable response rate of at least 75%. Survey research? (p. 133) and chpt. 7