1

Research Methods in Social Science

Name

Department, Institution

Course

Instructor

Date
 2

Research Methods in Social Science

Chapter 6 in the textbook describes different research methods and research feasibility.

The primary analyses revolve around the various research strategies. The strategies are described

as “a general approach to research dictated by the type of question that the research study hopes

to address”[ CITATION Gra15 \l 1033 ].  After discussing these strategies, the chapter moves on

to internal and external validity. There are five types of research methods to consider. These

include the descriptive, correlational, experimental, quasi-experimental, and non-experimental

research methods.

Descriptive research is a type of study that focuses on describing a group, condition, or

phenomena. It focuses on addressing the questions of how, when, when, and where instead of the

why of a research study[ CITATION Gra15 \l 1033 ]. This is due to the fact that it is important to

get a thorough view of a research topic before looking into why it exists in the first place.

Correlational research is a form of non-experimental analysis in which the researcher

analyses two variables and evaluates their statistical association (i.e., the correlation) with little

to no attempt to monitor extraneous variables. When researchers are unsure whether or not the

predictive interaction is causal, this approach is used (Gravetter & Forzano, 2015). Another

reason a researcher would use a correlational analysis is that the observational relationship of

significance is assumed to be causal, but modifying the independent variable is difficult,

unrealistic, or unethical.

Experimental research is used to assess not only whether two variables have a significant

relationship but also whether the relationship is causal and validated by statistical analysis. The

aim of the Experimental Testing Approach is to prove that variables have cause and effect
 3

relationships (Gravetter & Forzano, 2015). This is shown in an experiment by demonstrating that

adjustments in one variable cause changes in another variable.

The manipulation of an independent variable without the randomized assigning of

subjects to variables or orders of conditions is known as quasi-experimental research.

Nonequivalent group designs, pretest-posttest designs, and disrupted time-series designs are

among the most common[ CITATION Gra15 \l 1033 ]. Since it entails manipulating the

independent variable, quasi-experimental testing avoids the directionality issue. However, since

it does not require arbitrary assignment of situations, it does not eradicate the issue of

confounding variables.

Non-experimental research does not include the selection of an independent variable, the

random distribution of extraneous variables, or both[ CITATION Gra15 \l 1033 ].

Differences Between the Experimental,

Quasi-Experimental and Non-Experimental Research Strategies

There are several Differences between experimental, quasi-experimental, and non-

experimental research strategies[ CITATION Gra15 \l 1033 ]. Participants in experimental

studies are randomly distributed to either the treatment or control group, while in a quasi-

experiment, they are not. The control and treatment groups in a quasi-experiment are also

different since they undergo several aspects that are sometimes unclear or unknowable. As a

result, the researcher must try to compensate for as many of these variations as possible

statistically.

The distinctions between these research strategies lie in the degree to which the

researcher manipulates and controls the environment. In laboratory experiments, the researcher
 4

creates a controlled atmosphere and closely monitors the variables of interest. The non-

experimental analysis is conducted in a real-world environment in which the researcher is unable

to monitor all potential variables[ CITATION Gra15 \l 1033 ]. This means that non-experimental

experiments are more difficult to assess causality, but they are often more diverse and can cover

a wider variety of topics.

Give An Overview Of A Research Strategy Of Your Choice And Argue Why It Is

Important To Have One

The strategy of my choice the experimental research design mainly due to the benefits it

has over the other strategies. In experimental research, there are two sets of variables, one of

which is a constant and used to compare the differences between the two sets[ CITATION Gra15

\l 1033 ]. When there is not enough evidence to back up your conclusions, this analysis approach

comes in handy. First, you must create the truth. Experimental analysis collects the information

you need to make smarter decisions. Experimental techniques are used in all studies done under

scientifically appropriate conditions. However, researchers must agree that the shift in a variable

is primarily due to the manipulation of a constant variable for experimental experiments to be

effective. 

This is my favorite choice because I think it is the most detailed. It explores a cause-and-

effect relationship between the two variables. Other approaches either clarify the relationship or

fail to see any cause-and-effect interactions. I assume it is important to consider both the

interaction and the cause-and-effect in partnerships, as well as why this is occurring. It is

important to have a chosen testing approach such that a person becomes familiar with it and is

secure in using it to perform an experiment.

Explain How You Are Going To Ensure That Your Research Proposal Possesses Proper

External Validity

My analysis proposal would have enough external validation by ensuring that my

findings can be applied to a variety of “people, settings, periods, steps, and features other than

those used in the study.” This is possible because my study plan focuses on adults and youth in

particular professions, classes, and ages. Nonetheless, the experiment can quickly be repeated on

a different set of individuals of different experiences and attitudes. The impact of multiple

peaceful practices, such as reading, journaling, and nature walking, on stress control, can be

attributed to anybody at any moment.

In addition, my project will have proper internal validation, which is described as “factors

in the research analysis that raise doubts or concerns about the interpretation of the findings.” If I

use the experimental technique in my research, I would have to ensure that if any of my subjects’

stress levels are lower at the end of the trial, it is due to the peaceful practices and not to any

other causes. Other causes may include one of my research participants getting a demanding

career that quits his job in the middle of the study (while still engaging in peaceful activities),

and then his stress levels decline.

“Threats to external validity explain how one may be mistaken in generalizing from the

results of a single study is a challenge to external validity. When the influence of one factor (the

independent variable) is dependent on other variables, generalizability is usually constrained. As

a consequence, statistical interactions should be used to describe all risks to external validity.

One of these threats is Aptitude by treatment Interaction. It happens in a study with certain

characteristics that interfere with the independent variable, reducing generalizability.

Comparative psychotherapy experiments, for example, often use unique samples (volunteers,
 6

highly depressed, no comorbidity). Can psychotherapy be effective for non-volunteers,

moderately suicidal, or people with concurrent conditions if it is shown to be effective for these

study patients? Otherwise, the study’s external validity will be undermined.

Although listing threats to validity can help me stop making unfounded generalizations,

many of those threats can be disarmed or neutralized in a systemic way to allow external

validity. Experiments from one sample should be “re-processed” or “re-calibrated” in order to

avoid population variations to produce accurate generalizations in a second population where

experiments are not possible.

References
 7

Gravetter, F., & Forzano, L.-A. (2015). Chapter 6: Research Strategies and Validity Flashcards
Preview. In F. Gravetter, & L.-A. Forzano, Research Methods for the Behavioral
Sciences (pp. 159-169). Wadsworth Publishing.