Anecdotal evidence is first or second hand information about an individual's personal experience. Both positive and negative anecdotal evidence are unreliable as they do not provide tangible data that can be stratified, analyzed or scientifically interpreted. It does however offer an insight into how people think and feel and can often be the source, and lead to, research hypotheses.
Case studies as a research design offer a better quality of data than anecdotal evidence as the data provides detailed information about an individual rather than just being causal observations. Case study data is collected in a more systematic manner than anecdotal evidence but is weakened by the fact that data it is not derived from direct observations of behavior but from interpretations of what the research individual or participants report about the behavior. Case studies often take a great deal of time; they are limited in scope and cannot answer questions about behavior and whether 'a' caused 'b'. Case study data is collected in a more systematic manner than anecdotal evidence but still lacks scientific control. Case Study data looks at individuals only; it cannot be generalized beyond that individual.
Descriptive studies, unlike case studies and anecdotal evidence, use a larger quantity of data that can be statistically analyzed. The quantity of data collected reduces the risk of small data samples distorting the research findings. The data collected from a descriptive study is analyzed to determine variables across different members of the groups. Descriptive data can be used to suggest a hypothesis for a particular phenomenon across the data but it cannot test the hypotheses. With descriptive studies we are unable to determine if our findings are a result of the variables we have identified. To do this we need to perform scientific experiments. Experimental studies, unlike descriptive studies, allow us to study the effects of independent variables on dependent variables within our research groups.
There are two types of experimental studies; between-subject designs and within-subject designs. Each of these designs use two types of variables. Independent variables are a subject's experience (the thing we manipulate) and dependent variables are a subject's behavior (the thing we measure). Between-subject designs look at two or more groups of participants where the independent variable is different across the groups. Within-subject designs look at an individual's behavior before, during and maybe after the experiment. The effects of the independent variable are judged differently within these two research designs. The effects of the independent variable in in-between-subject research are statistically analyzed whereas within-subject experiments look at the effect of the independent variable by noting changes in the participant's behavior. So, for between group experiments we are asking questions about a population that we want to infer to. In single subject experiments we are asking questions only about that individual.
Because anecdotal evidence, case study evidence and descriptive studies cannot explain why phenomenons occur it is essential to use experiments to study behavior.
When working with a subject, an animal, and we need to test a hypotheses regarding fundamental principles of behavior we need strong control so we can take that experiment into a laboratory to test a hypotheses. When looking at animal behavior and the consequences maintaining the behavior we need to study that in the animals' natural environment so it is not contrived. We choose the environment based on the questions we wish to answer. Experimental design allows a control of the data that we can scientifically interpret. This interpretation is not available with data obtained from anecdotal, case study or descriptive design.
Chance, P. (2008) Learning and Behavior, Wadsworth Cengage Learning
Article Source: https://EzineArticles.com/expert/Niki_Tudge/229016