Chapter 3 Summary - Mayer & Clark

Summary Chapter 3

Chapter 3 is one of the three foundational chapters in the textbook. The chapter focuses on the importance of evidence-based practice as it relates to e-Learning courses.  Evidence-Based Practice addresses the importance of doing research on a topic that will be taught to a class.  Before a lesson can be prepared  the instructor should "look at what the research has to say (Clark & Mayer 2011) research should be done to find cases where the particular lesson's subject has been taught. The chapter outlines three approaches to research on instructional effectiveness.  The main point of instruction is to be effective and to teach the learner, there are three essential questions that should be asked about the research, what works, when does it work and how does it work.  The research will answer at least one of these questions to determine the overall instructional effectiveness with the key focus being on whatever works while presenting evidence that outlines how it works and under what conditions.

Experimental comparisons must include criteria that relates to the lesson that the instructor is preparing. If the comparative research does not apply to the lesson then it has no relevancy and therefore the research is virtually useless according to chapter 3.  The research studies that are selected should be close to the lesson that is being prepared and focus on the type of instructional method and learning environment that reflects yours, if you are developing an online course, for example, the research should include e-Learners. The chapter encourages experimental comparisons as the criteria for determining whether or not the research is good.  The last point that relates to good experimental comparisons include experimental control, random assignment and appropriate measures (Mayer, 2011a).

In some studies, there is no difference that shows up in the control group and the experimental group.  There are several reasons why this could occur.  The first reason could be as simple as the treatment does not actually affect the students.  Some studies could show no difference if the study size is not big enough.  The actual test that was looking for significance might not be adequate to detect a difference. Another reason could be that the treatment was not different enough to actually make a difference in the groups. Also, if the material was easy for the groups being tested, the treatment would not be effective because there would be no need for a treatment. Researchers should be careful about confounding variables.  An example of a confounding variable would be if there were two groups and one group was stacked with learners with higher intelligence scores.  That would make a difference in the outcome. When comparing two groups, there is a control group which does not receive the treatment and the test group which does receive the treatment. To find the standard deviation of these two groups, look at the test scores and see how they are spread out. The variation of these group scores shows the standard deviation. Two statistical measures that this chapter focuses on are probability and effects size.  In looking for treatments that are considered effective, probability should be less than .05 (p<.05) and the effect size should be .5 or greater.

When trying to decide if a research is relevant, look at who the learners are. Do the learners in the research design match the learners in the experiment you are designing or are they transferable?  Make sure the study is an experimental study that has a random and control group. If the research has been replicated and results have proven to be the same, that is a good indication that it was a good research design.  If the test measures learning recall rather than application then it might not work for workforce learning goals that are application based. Always go to the results section and look to see if there is a significance score (p<.05) and an effect size of .5 or greater. One special type of research article is a meta-analysis.  A meta-analysis looks at experiment results on many studies that test the same effectiveness of the same instructional method and record the effect size for each study and compute an average effect size from across all of the studies. 

Reflection

As a professional, I should choose an instructional strategy because of its proven worth and not because of its appeal to me as the instructor.  There is a plethora of scholarly research helping us practitioners make sound decisions.  We have to read that literature and be prepared to utilize effective practices and do away with ineffective practices.  For about a decade now, we have all been inundated with “Best Practices” in reading, writing, math, science, technology…the list goes on and on.  Since this world of eLearning is new to us as possible instructional distance designers, we need to familiarize ourselves with what is out there so we can capitalize on the time that we have with our distance students. Before this class, I would have designed a course just from the courses that I have already taken here at Liberty and also from East Carolina University in Greenville, NC.  This chapter is another reminder that research does matter. It is an excellent brief summary of what we should look for in research.  I am not sure if I will become an instructional designer when I graduate.  I do know that I want to teach online in the near future.  It will cause me to question the instructional designs and to make sure they are research based.  I will not be coming into the course blindly.  I will also know where to look for answers

Reference

Mayer, R. & Clark, R (2011). e-Learning and the science of instruction: Proven guidelines for consumers and designers of multimedia learning. San Francisco, CA: John Wiley & Sons, Inc.