feature

The Science Teacher—February 2020 (Volume 87, Issue 6)

By Scott B. Watson and Michelle J. Barthlow

Ms. Jones stared at the stack of biology quizzes and wondered what went wrong. She thought about the lesson plans and wondered what she should have done differently. Ms. Jones is not alone in wondering how to improve student learning and classroom instruction. To improve student achievement, educators must determine what is working and what is not. This article focuses on basic research skills that teachers can utilize to conduct studies in classrooms and schools for the purpose of determining the effectiveness of instructional techniques and curricula. While the emphasis is on quantitative research, the basics of qualitative research are mentioned as well.

The purpose of educational research in general is to develop information which can be used to improve education. Alexakos (2015) stated that teachers conduct research to answer these questions about their own practice: How am I doing? How can I improve? What works? To answer these questions, teachers may conduct action research.

Action research is a special form of educational research. Gall, Gall, and Borg (2007), authors of some of the most respected educational research texts, define action research as: “A type of applied research the purpose of which is the improvement of education professional’s own practice.” Lesha (2014) describes action research as being a cyclical or spiral process that begins with a teacher-researcher identifying a problem, investigating the problem, taking action, evaluating the results of the action, and then repeating the process. In doing so, teachers can develop the most appropriate strategies for their own classroom or school.

Action research is not necessarily very different from other forms of educational research. The main difference is that it is conducted by practitioners in the schools instead of someone from outside the school, such as a university professor or another researcher. With schools focused on learning outcomes for students and the call for decisions based on student data, teachers need the skills and confidence to scientifically evaluate their own practice in order to make curriculum and instructional decisions. Action research provides teachers with the data needed to make informed decisions to benefit their students and improve their own classroom practice.

Action research is a great way for teachers to experience the 3D (three dimensional) approach of the *Next Generation Science Standards* (*NGSS*). *NGSS* incorporates the three dimensions of learning science: crosscutting concepts, science and engineering practices, and core ideas. As teachers embrace the *NGSS*, they will consider the crosscutting concepts of effective teaching and focus on the core ideas to be taught as they plan for instruction. Through action research, teachers can investigate their natural world—classroom instruction—to determine what is and is not resulting in learning gains for their students.

So why should science teachers be interested in research? Although most educational research is conducted by college professors and other professional researchers, teachers can enhance their own knowledge and may contribute to the research base through research in their own classrooms (Abell 2007). Science teachers, because of the nature of their discipline, have a natural interest in research, and often have a good understanding of research methods.

Teachers seem to know many things intuitively (and through experience). A good example is using a hands-on approach to teaching science. The idea is that simply doing many activities is conducive to learning, which is not necessarily the case. Research findings indicate that if students do not fully understand what the activity is all about, very little learning really occurs (Gough 1990; Nadelson 2009). In order to maximize learning (and achievement), a minds-on approach should be added to the hands-on approach. This should include using higher-order thinking and problem-solving skills in addition to simply participating in an activity (Lumpe and Oliver 1991).

As another example of the practicality of conducting and understanding research for science teachers, consider the case of Mr. Nolan, a young chemistry teacher. Almost all of Mr. Nolan’s classes in college included midterm and final exams, mostly using a multiple-choice format. When he started teaching, he followed the same model with his own students. His whole evaluation system was based on teacher-made tests.

In an effort to increase his own knowledge in science and in education, Mr. Nolan enrolled in a graduate degree program at his local university. One of the early courses he completed was in educational research. During that time, he learned about reliability (internal consistency) of tests. Reliability is normally determined on a scale of 0 to 1, with 1 being perfect. It is an indicator of the precision, consistency, and stability of an instrument (Gall, Gall, and Borg 2007, p. 149).

One of his projects was to determine the reliability of one of his own tests. He picked one that he considered his best, and he ran a reliability figure using one of many available computer programs. Much to his dismay, he found that the reliability of his prized, multiple-choice measure was a .58, which is substantially lower than what is minimally acceptable for research purposes. He realized that relying on imperfect tests alone for grading purposes was a mistake. This discovery changed his teaching almost immediately, and he started including more projects, presentations, and practical labs as part of his assessment system. This also produced greater interest and participation among his students.

Two types of research methods are *qualitative* and *quantitative* studies. Denzin and Lincoln (1994) describe qualitative research as “interpretive, naturalistic….Qualitative researchers study things in their natural setting, attempting to make sense of, or interpret, phenomena in terms of the meanings people bring to them.” Quantitative research “…describes and explains…reality by collecting numerical data on observable behaviors…and by subjecting these data to statistical analysis” (Gall, Gall, and Borg 2007).

Qualitative research differs from quantitative research in that it depends on numerical data; no statistics are needed. Qualitative research methods include interviews, surveys, and observations. Teachers could prepare a questionnaire to determine what methods and activities students feel are most beneficial to their learning. For more in-depth information, a teacher could conduct interviews with students.

A focus group of students can provide valuable insight into their experience in the classroom. Ary, Jacobs, and Sorenson (2010) point out that an advantage of a focus group is that participants respond not only to the interviewer but also to each other. These student-to-student interactions can result in more information than is typically collected in a one-on-one interview or survey.

Quantitative research is the systematic study of the relationships among variables. A variable is anything that can change during a study. An independent variable is sometimes referred to as the manipulated variable as it is deliberately changed (manipulated) during an experiment. A dependent, or responding, variable is one that may change as a result of the experiment. A controlled variable is a variable that you try to keep constant during the experiment. An extraneous variable is an outside or unknown variable that you have no control over.

For an action research project, Ms. Jones is curious about using the flipped classroom model, in which students first watch instructional videos outside of class and do homework and practice problems in class (Brunsell and Horejsi 2013). For one unit of study, she decides to teach half of her class periods using the flipped model and the other half using the traditional model, where students engage in learning activities in class and do homework at home. She will give the same unit assessment to each group and compare the results.

Her independent variable is classroom pedagogy (flipped classroom vs. traditional). Her dependent variables are unit test results. The controlled variables include the length of time for the unit, the state standards, and the homework practice problems utilized. The examples given in class and on the teaching videos are all the same. Ms. Jones will compare the two group’s performance on the unit test using a *t-*test, which will allow her to determine if one group performs significantly better than the other on the assessment based on her pedagogy. Knowing what works best for students in her classroom will allow Ms. Jones to improve her teaching skills and will likely increase learning for her students.

Simple forms of correlation research can be used to determine if there is a relationship between two continuous variables. A continuous variable is one that has a maximum value and a minimum value and can be any value in between (Gall, Gall, and Borg 2007). A correlation will not show causation but will show if a relationship exists between two variables. An example action research project would be to determine if using a web-based practice quiz site prepares students for tests. A correlation could be performed to determine if the number of practice problems a student answers correctly correlates with their summative test score.

Quasi-experimental designs are used where random assignment of individuals to experimental and control groups is not possible. This is what you will use if you are interested in determining the effectiveness of a strategy used with whole-class instruction. The previous example of Ms. Jones investigating the effectiveness of a flipped classroom is an example of a quasi-experimental design. She could not randomly assign the students to each treatment group, so she used intact classes for each treatment.

Causal-comparative research designs work the same way as quasi-experiments, except that the effect has already occurred (*ex-post facto*). Since the effect has already occurred, the researcher cannot directly manipulate the variables. Instead, they will look at what the results were based on and the conditions at the time. An example of a causal-comparative study would be the difference in test scores based on biological sex.

Quantitative studies can become rather complex because of the statistical tests that are often involved. Even so, simple statistics such as correlations and *t*-tests are not too difficult for many science teachers.

Many statistical tests are available at no cost online or using Microsoft Excel. Some can even be performed with scientific or advanced calculators. Statistical analysis sources may be found by doing simple internet searches. One example of a free online resource for performing statistical calculations is GraphPad.

So where do you start in conducting action research? As with any form of research, you begin with a problem or question. Once you make a firm decision of what to investigate, you do background research and design your study. It always helps to have a good, educational research text on hand when designing a study or making decisions on topics to investigate.

Teachers who gain expertise in research methods often help others begin their own projects. The results of action research studies may then be presented at professional conferences and possibly even published in professional journals.

Science teachers, due to their content training, have a real head start on most teachers in other fields when it comes to conducting research. Science teachers also have an advantage in their understanding of research because they already know that research can be fun, so get started!

Literacy Research Teaching Strategies

Literacy-Research-Teaching Strategies