There are 3 major types of research design; Quantitative Research Designs, Qualitative Research

Designs, and Mixed Research Designs. This summary will focus on Quantitative Research
Designs. Quantitative Research Designs consists of Experimental Research Design and
Nonexperimental Research Design.

EXPERIMENTAL RESEARCH DESIGN

In Experimental research designs, (at least one) independent variable is manipulated to control
what the participants will experience in the study. Those participants will be divided into 2
groups: experimental/nonexperimental group or control group.

-The independent variable: treatment/experimental variable

-The dependent variable: effect/posttest variable (the result of the independent variable)

Experimental research designs normally compare 2 groups. It has 3 possibilities:

 Compare 2 different approaches

Example: A researcher wants to compare the effects of a project-based method to enhance

 Compare a new approach and the existing approach

Example: Compare the effectiveness of a student-centered approach (new approach) to a

 Compare the different amounts of the same approach

Example: The effect of assigning homework 30 minutes per day on students' academic
achievements to the effect of assigning homework 60 minutes per day.
There are 3 major types of Experimental Research Design:

1. Pre-experimental Designs

The pre-experimental designs do not use control groups and are conducted before the researchers
conduct the true experimental designs. The purpose of this design is to analyze and explain in
detail a group of participants.

There are 3 types of pre-experimental designs:

One-shot Group Case Study

In this design, a single group is studied after got some treatment. The example is a study on the
effects of a flipped classroom method on students' speaking skills by observing the performance
of a single class after the flipped classroom method has been applied.

One-Group Pretest-Posttest Design

In this design, a single group is observed twice, before (pretest) and after (posttest)
treatment/intervention).

Static-Group Comparison Design

This design uses 2 groups, but only 1 group receives the treatment. This design involves analysis
statistics, which is an independent t-test.

2. True Experimental Designs

In this design, each participant is randomly selected, and then are randomly assigned to any
group or to different treatment conditions (experimental/control). The use of a random
assignment is to provide an equal change to each participant to be assigned before treatment.
Random assignment is the unique characteristic that only exists in true experimental designs.

True Experimental Designs are divided into 3 types:

Posttest-Only Control Group Design

This design does not use pretests but use posttests. The posttest results used to evaluate the
effectiveness of a treatment on a dependent variable.
 Pretest-Posttest Control Group Design

In this design, both groups (experimental and control) receive pretests before treatment and
posttests at the end of the study, but each group get a different treatment.

Solomon Four-Group Design

This design combines the previous designs which are the posttest-only control group design and
pretest-posttest control group design. As its name suggests, this design has 4 groups, 2 groups
receive pretests, and 2 are not. Each experimental and control group consists of 1 pretested group
and 1 not pretested group. The advantage of this design is its ability to figure out whether an
interaction between pretest and treatment causes a change in the dependent variable.

3. Quasi-Experimental Design

The participants in both the experimental and control group are not randomly assigned in this
design. The quasi-experimental design is used when true experimental research, which uses
random assignment, is not possible to conduct.

The quasi-experimental design has 3 forms:

Nonrandomized Control Pretest-Posttest Design/Nonequivalent Control Group

Instead of assigning individuals or participants, this design assigns entire groups to different
treatments. In this design, the researchers need to ensure that the groups are as equivalent or as
similar as possible to reduce validity threats.

Time Series Design

This design is a group version of the one-group pretest-posttest design. In this design, a single
group is first given a pretest multiple times, and then receive a treatment, and after that, the
group get a pretest again. Pretests are given several times until the researcher obtains stabil
scores. This design allows researchers to measure the results before and after a treatment has
been implemented and helps them to analyze changes in behaviors or variables.

Counterbalanced Design
In this design, all groups receive the same treatments, but they get different orders. For example,
there are 4 groups (group 1, 2, 3, and 4), with 4 treatments (A, B, C, and D). Group 1 might
receive the treatment A first, and then B, C, and lastly, D. Group 2 can receive treatment B first,
then treatment C, D, and A. And so on.

After each treatment, each group is given posttests (treatment-posttest-treatment-posttest, and so

NONEXPERIMENTAL RESEARCH DESIGNS

Nonexperimental designs are the opposite of the experimental designs. They do not use random
assignment and do not manipulate the independent variables. The nonexperimental designs are
divided into 4 types:

1. Correlation Designs

The purpose of this design is to analyze whether the relationship exists between two or more
variables. For example, motivation and academic achievements are believed to have relationship.
The students with great academic achievements typically have high motivation to study, and vice
versa. There should be at least 30 participants in this design.

2. Survey Designs

Survey designs are based on the survey or the process of gathering information from a
population on the phenomenon or issue. Survey designs have 2 forms; Cross-Sectional Surveys
and Longitudinal Surveys.

 Cross-Sectional Surveys

In this design, the information is gathered from selected participants at a specific time. The
cross-sectional surveys provide insight into the population's current situation. They are also
quick to conduct.

 Longitudinal Surveys
Longitudinal surveys assess changes and developments by gathering the data repeatedly (more
than one time).

3. Causal-Comparative Design

Causal-comparative research or known as ex post facto research which is a research design used
to identify whether the relationship exists between the dependent and independent variable by
comparing 2 or more groups (which are grouped into experimental and control
groups/comparison groups).