Understanding the Science Model
The following excerpt from Life Journey Through Autism: A Parent’s Guide To Research is provided with the kind permission of The Organization for Autism Research.
UNDERSTANDING THE SCIENCE MODEL
Reading and Interpreting Research
Reading and interpreting research can be a difficult task that will require new skills for the beginner. However, with perseverance and practice, you can gain this expertise and benefit from it. This section is written to give you a basic understanding of the science model before you get started reading the research.
What Can Research Do?
There are two kinds of research: Basic Research and Applied Research. Both kinds of research are necessary to advance science and apply science to the benefit of society.
What is the difference between basic and applied research?
Basic research is designed to answer questions about a fundamental or basic problem for the purpose of acquiring knowledge. Biomedical research is often an example of basic research. Research in other scientific areas may overlap between basic and applied research. For example, a researcher may want to answer a question about the relationship between language development and school readiness in children with autism, and then design a basic research study that collects information about this phenomenon.
Applied research is concerned with applying the knowledge gained through basic research to practical, everyday problems. For example, through basic research, we might learn from the example above that children having autism with verbal skills need less support in an academic setting. Using this information gained through basic research, an applied researcher might design a study to see if a particular program designed to increase verbal skills in preschool children with autism increases their readiness for school.
Where are these two types of research conducted?
Basic research is usually conducted in laboratory or clinical settings, such as university and medical laboratories. Applied research is usually conducted in everyday settings, such as the home, schools, and other community settings.
Studying a Study
A research study is designed to investigate a topic of interest. A study begins with a research question, such as: “How does the play behavior of children with autism differ from the play behavior of typically developing children?” Once a research question has been posed, the researcher designs a study to answer the research question.
Five Basic Types of Research Studies
There are five basic types of research studies; each type of study can answer different kinds of questions. The five types (A−E) of research studies are discussed below. An example of each type of study is provided in the "Sample Research Articles" section.
A. Case Study
A case study looks at a single subject (e.g., a child with autism) or a single case, (e.g., a classroom for children with autism). Single-subject studies are typically used in research in therapeutic settings where there is a one-on-one relationship, such as in counseling. Many studies with children with autism, particularly those looking at behavioral treatments, use a single-subject design (a subset of which falls into a specialized category referred to as a "single-subject experimental design").
The advantage of a case study is that a lot of information can be gathered about one person or one case.
The disadvantage of a case study is that the information gathered about one person or a small group of people cannot be readily generalized to other people or other cases in other situations or other settings.
B. Correlational Studies
Correlational studies look for a relationship between two or more variables, or things that naturally occur in the same environment. Correlational studies cannot tell us anything about cause and effect, only that there is a relationship between two or more things. For example, a study might be designed to determine if there is a relationship between the number of children with autism in a particular community and month of birth. Figure 1 presents a bar graph of a study of number of children with autism and month of birth. The graph illustrates that more children with autism were born in the winter months (November, December, and January) than in the summer, spring, or fall.
Although this graph shows a relationship between cases of autism in one community and month of birth, we cannot make the assumption that a winter birthday causes autism. It may be the case that some other variable that we have not considered (such as illness in the mother) is having an effect on the two variables under consideration.
The advantage of correlational studies is that a lot of information about a large number of people can be gathered at one point in time.
The disadvantage of correlational studies is that it is not possible to control other factors outside of the study that might influence the research.
C. Longitudinal Studies
Longitudinal studies can give us information about how people develop over time. These types of studies follow one group of people (referred to as a cohort) across time, measuring the same behavior multiple times. For example, we may want to determine if children with autism do better on performance tests in self-contained or inclusive classrooms. The best research design for answering this question would be to follow one cohort of children with autism who have spent time in both environments. In Figure 2, this design is presented with a line graph.
As illustrated in the graph, children in this cohort performed better in grades 1 and 4 when they were in inclusive classrooms than in grades 2 and 3 when they were in self-contained environments.
The advantage of longitudinal studies is that they allow researchers to pinpoint times during development when changes occur.
The disadvantage is that longitudinal studies require a long time to complete. (Note in the example in Figure 2, the study took 4 years to complete).
D. Experimental Studies
Experimental studies are used to study cause and effect. Experiments are controlled so that one variable is manipulated by the researcher to determine its effect on other variables.
In experimental designs, participants are randomly assigned to receive either the intervention (the experimental group) or to a control group. The control group completes all the same steps as the experimental group, except the intervention that is under investigation. Therefore, if the study is well controlled, the differences between the experimental and control groups at the end of the study will be due to the intervention.
A common type of experimental study is an intervention study testing the effectiveness of a treatment program on outcomes such as performance scores.
Using this example, the variable that is manipulated by the researcher is the “educational program.” In this study, subjects in the experimental group would receive the new program, while subjects in the control group would receive a standard program.
The purpose of the study would be to determine if the test scores of students receiving the new method would be higher than the scores of the students receiving the standard method.
The advantage of randomized experimental studies is that the research design allows researchers to examine cause and effect relationships.
The disadvantage of a randomized experimental study is that the results cannot always be generalized to the real world. This is because this kind of study lacks “real-world” authenticity; that is, what occurs in a controlled environment of a study may be very different from what might occur in a real-life setting.
A cornerstone of experimental research is the concept of randomization. Participants in the experimental study are randomly assigned from the population to either the experimental or the control group. This assignment is done randomly so that the groups will be exactly the same. Only if the groups are exactly the same can the researcher determine if the differences between them at the end of the study are due to the intervention. Figure 3 illustrates the randomized experimental method of research.
E. Clinical Trial StudiesClinical trial studies are one type of randomized experimental studies. Clinical trial studies are most likely to be conducted in medical or other clinical settings. Similar to experimental studies, clinical trial studies employ an experimental/control group protocol, in which subjects are randomly assigned to receive the experimental treatment (for example, a medication to treat autism symptoms) or a placebo (a sugar pill).
In both randomized experimental and clinical trial studies, participants are not informed about their group assignment to guard against participant expectations about intervention or treatment. In addition, in well-designed studies, experimenters are not informed about group assignment to guard against researcher expectations about intervention or treatment. That is, if either the participant or the researcher expects to get better because they know they are getting treatment, they may actually get better, whether or not the treatment is actually effective! This uninformed state is what researchers refer to as being blind to the conditions of the experiment. When both participants and experimenters are uninformed, it is referred to as a double-blind study.
The advantage of clinical trial studies is that they can determine the effectiveness of an intervention. This is because the intervention is compared with a placebo or control condition.
The disadvantage of clinical trial studies is the same as randomized experimental studies; that is, the results of the study cannot always be generalized to a real life setting.