Research designs can be broadly classified into two categories, namely quasi experimental and experimental.
Statistics Solutions is the country’s leader in research design and dissertation consulting. Use the calendar below to schedule a free 30-minute consultation.
Aligning theoretical framework, gathering articles, synthesizing gaps, articulating a clear methodology and data plan, and writing about the theoretical and practical implications of your research are part of our comprehensive dissertation editing services.
The research designs are said to be quasi experimental only if the subjects are randomly assigned to the groups and the statistical controls are used by the researcher.
Non equivalent control group research designs are one of the quasi experimental designs. Cook and Campbell (1979) gave some non equivalent control group designs.
One group posttest only research designs are also sometimes called one shot case studies. This is one of the non equivalent control group research designs. This type lacks a pretest baseline, thus, it results in making invalid conclusions.
Posttest only research designs with non equivalent comparison group designs are a kind of social science research design. In this kind it is quite impossible for the researcher to draw valid conclusions about the treatment effects which are entirely based on posttest information.
Posttest only research designs that predict higher order interactions are used in the cases when the expectation of the treatment effect interrelates with the third variable. However, these types are confined to the possible challenges of validity due to certain factors.
One group pretest-posttest research designs are common but defective in social science. These research designs are also known as proxy pretest-posttest designs.
Two group pretest-posttest research designs using an untreated control group are a kind of classic experimental design.
Double pretest research designs help in strengthening the pretest and posttest designs. This kind can be established only if there exists a particular trend in the data that is independent of the treatment effect and is measured by pretest.
Interrupted time series research designs are one of the quasi experimental designs. Cook and Campbell (1979) list certain time series designs.
Simple interrupted time series research designs are the expansion of one group pretest-posttest designs into multiple pretests and posttests. These one group pretest-posttest research designs do not have the control group and therefore make it difficult for the researcher to assess other confounding factors.
Interrupted time series research designs with removed treatments are powerful. These are powerful because in these designs, the threat of certain unwanted factors are removed.
Interrupted time series research designs with multiple replications are simply an interrupted time series design with removed treatments, except that the treatment and the removal in these occurs multiple times.
Interrupted time series research designs with switching replications require a much higher level of control over the subjects. But is stronger in ruling out the threats of invalid conclusions.
Interrupted time series research designs with non equivalent dependent variables have the goal of obtaining the dependent variables that are related to the dependents being studied. In these designs, the related variables are not assumed to be correlated with the treatment variables.
A non experimental research design is not a kind of quasi experimental design because these types do not use statistical controls.
The research designs are said to be non experimental only if there exists a systematic collection of the data with respect to interest of study that are not considered experimental (as there are no control groups or randomization of the subjects).
Some qualitative approaches are applied typically to such research designs. These include approaches like case study designs, content analysis, participant observation, etc.