Unit II Research Discussion Board

RCH 8301, Quantitative Research Methods 1

Course Learning Outcomes for Unit II Upon completion of this unit, students should be able to:

1. Differentiate between quantitative, qualitative, and mixed methods research design methods. 1.1 Analyze descriptive, comparative, and associational research approaches.

2. Distinguish between experimental, quasi-experimental, and nonexperimental research designs.

2.1 Discuss the similarities and differences between experimental, quasi-experimental, and nonexperimental research methods.

Course/Unit Learning Outcomes

Learning Activity

1.1

Unit Lesson Chapter 4, pp. 56–66 Chapter 5, pp. 69–85 Chapter 6, pp. 89–102 Chapter 7, pp. 106–117 Chapter 8, pp. 120–130 Unit II Essay

2.1 Unit II Essay

Required Unit Resources Chapter 4: Research Approaches, pp. 56–66 Chapter 5: Randomized Experimental and Quasi-Experimental Designs, pp. 69–85 Chapter 6: Single-Subject Designs, pp. 89–102 Chapter 7: Nonexperimental Approaches/Designs, pp. 106–117 Chapter 8: Internal Validity, pp. 120–130

Unit Lesson

Research Approaches

In this unit, we will learn that research is conducted to analyze a problem and that research is mainly deployed to clarify a generalized assumption. Research tries to explain the problem and tries to identify its solution. Theories and observations are the pillars of research, and they strengthen each other. From an observation, a theory can be inducted, and from a theory, observation can be deduced. Research is done to unravel things that were hidden earlier, and research methods are used to shape how research is created, how it is conducted, and how those findings can be theorized. Research methods has become a field of academic inquiry, and, over the years, it has created a body of knowledge and approaches to constructing, conducting, and theorizing research. The research approaches can be broadly separated into the 1) randomized experimental approach and the 2) quasi-experimental approach.

UNIT II STUDY GUIDE

Quantitative Research Approaches, Questions, and Designs

RCH 8301, Quantitative Research Methods 2

UNIT x STUDY GUIDE

Title

Randomized Experimental Approach

The randomized experimental approach is the most rigorous form of experimental research. It prevents selection bias and ensures against accidental bias. Randomized experimental research is considered the gold standard against which all other types of research are compared. Experimental research is done either to accept or reject a hypothesis, and these experiments are conducted in order to answer a generalized assumption and to check for its validity. The aim of this kind of research is to find the cause-and-effect relationship, and it is primarily carried out to understand the causal relationship. Researchers manipulate the variable to a control group and will note the changes (if any happens). They note such changes in order:

1. to identify the causal relationship (the effect happens only when the cause happens); 2. to verify the consistency of the cause and effect (whenever the cause changes, it also changes the

effect); and 3. to quantify the relationship between the change and the effect.

Before starting an experiment, the researcher should visualize what he or she needs. The researcher must note the observations that were deduced from the earlier theories. The gaps must be identified in the research and addressed. If the problem is properly designed, it is halfway to solution. Two sets of the people who are selected from similar backgrounds are formed groups. The group is made in such a way that one group is equivalent to another. This experiment happens in a controlled environment. In the two groups, there is a control group in which the settings of the groups are manipulated, and another group is denied what the first group gets. Now the difference in them is noted. In this stage, the obtained data is analyzed, and a conclusion is drawn from it. The conclusion is studied with the observation of the similarities and the deviations—both of which are noted. The data here is known as the raw data. The important thing about research is its transparency and consistency. The data of the research should be made public so that anyone interested can cross-check and validate the findings.

Research Designs

In the previous unit, there was an initial explanation of variables. We learned that there are two kinds of variables: independent variables and dependent variables. Independent variables can be explained as the cause, while dependent variables are the effect of that cause. When there is manipulation in the independent variable, there is a change in the dependent variable.

The independent variable can be further divided into an active independent variable or an attribute independent variable (Gliner et al., 2017). A variable that can be manipulated by the researcher is known as the active independent variable. In a controlled experiment, this variable can be manipulated. For example, in a dieting experiment, if the researcher changes the dieting regimen, then that variable is known as the active independent variable. A variable that cannot be manipulated by the researcher is known as the attribute independent variable. Even in a controlled experiment, certain characteristics are beyond the researcher’s ability to manipulate, and such characteristics describe an attribute independent variable. For example, the participant’s age, race, and mental rigor fall beyond the purview of the researcher’s manipulation, and such characteristics describe an attribute independent variable. As human nature can be frail, there is always an inherent bias in selection. In order to reduce this bias, the randomization method is used to select the sampling, and human intervention in selecting the participant should be kept to a minimum. Randomization enhances the conclusion of the experiment overall.

(Plantfelicity, n.d.)

RCH 8301, Quantitative Research Methods 3

UNIT x STUDY GUIDE

Title

Quasi-Experimental Design

There are many areas where the randomized experimental designs cannot be used. To study such conditions, quasi-experiments are considered a suitable alternate. In the researcher’s world, this model is unscientific and unreliable as it does not carry the same rigorous standards of the experimental design. Nevertheless, it can be used to measure huge sizes such as cross-cultural behaviors or public policy effectiveness in similar communities. In this design, the researchers can only observe the causal variable but are not able to manipulate it. These types of research approaches are more realistic than the experimental designs as they reflect the real world. Reactions of the test subjects are more realistic, and the outcomes can be closer to the answers. These experiments reinforce the experiments in a controlled environment and can be used to increase its reliability. However, they are more prone to human error. The bias is much more prominent than in the controlled experiments (Price et al., 2015).

Nonequivalent Group Design

Nonequivalent group research design is one of the most frequently used in research where the researcher does not assign these types of groups. When participants are not selected and controlled, the groups tend to be more dissimilar. Nonequivalent groups can be defined as the groups where the participants are selected nonrandomly and post-intervention studies are conducted. We try to select groups that are as similar as possible so that we can fairly compare the treated one with the comparison one (e.g., average educational qualifications across various cultures, effectiveness of public policies across similar income countries). In nonequivalent group design, only one group is exposed to the treatment, and the other is not. When a new intervention is made in an ongoing project, this method can be used to gauge the effectiveness of the program. For example, a posttest intervention can be made to assess a new teaching method in a school. Its outcome can be noted. If similar groups are selected for the experiment, it will increase the reliability of the experiment and will reduce the confounding variables. To increase the reliability and credibility of the posttest design, pretest and posttest intervention can be introduced. A treatment group is given pretest and posttest intervention before and after the introduction of the changes. The nontreatment group where no such change applies also can be given this intervention to measure how effective the change is. For example, you could introduce a nutritional intervention scheme in a school. The pretest and posttest design can bring out how the nutrition of the child improves the learning outcomes of the child. The time series research design takes measurements over a series of time. Data is gathered periodically in a timely order, and the data consistency is checked. Time series design is relevant across various research and observational studies, and the long-term deviation, impact, and change of a phenomenon can be measured. Time series is a repeated periodical measurement that is taken across equally spaced time intervals. For example, the level of rain across a region through various time periods can strengthen the climatic models. The public policy’s health can be measured by analyzing its effectiveness. The data gathered when the policy was introduced (i.e., how the landscape has changed in the midcourse and its lasting impact) can be measured through the time series model.

Nonexperimental Research Design

Nonexperimental research is normally descriptive or correlational since variables cannot be manipulated, and participants cannot be assigned randomly; therefore, no causal nature can be determined. The nature of the research question determines whether the researcher chooses between an experimental and nonexperimental research design. If the research question involves a causal relationship and an independent variable that can be manipulated, the experimental approach is typically preferred. However, a nonexperimental research design is preferred when there is a single variable being studied, when there is not an active independent variable, or when there is a non-causal statistical relationship between variables. Although nonexperimental research can be both quantitative and qualitative, our focus will be on the three quantitative nonexperimental research approaches: descriptive, associational, and comparative. Many

RCH 8301, Quantitative Research Methods 4

UNIT x STUDY GUIDE

Title

research studies contain more than one hypothesis, so many use a combination of research approaches to construct the hypotheses.

Internal validity

One factor in determining the quality of a research study is its internal validity. Validity is a tool to measure whether the intention of the research matches the outcome of the research, and it measures how well an experiment is done. It shows the confidence that can be placed in the cause-and- effect relationship in the research study. Internal validity shows us how well a study is organized and executed to prevent errors and bias. The internal validity’s scope is within the research only and focuses on these issues:

 How well was the research done?

 Does it have confounding variables? Internal validity and confounding variables share an inversely proportional relationship. When the confounding

variables increasingly affect the outcome, the internal validity will be less. The higher the internal validity, the higher the credibility of the research will be (Trochim, 2006). There are a number of potential threats to internal validity. Threats, such as maturation (i.e., when the participants in the study change over time) and history (i.e., when an event occurs beyond the researcher’s control), affect the outcome of a study. Another influencing factor is testing, which is when participants do better on the posttest than on the pretest due to the experience gained. Instrument occurs when there is inconsistency between the last interview and the first one (Gliner et al., 2017). Therefore, it is important to have a well-designed research study with high internal validity.

References

Gliner, J. A., Morgan, G. A., & Leech, N. L. (2017). Research methods in applied settings: An integrated approach to design and analysis (3rd ed.). Routledge.

Hafakot. (n.d.). Data validity concept (ID 88877337) [Illustration]. Dreamstime.

https://www.dreamstime.com/stock-illustration-data-validity-concept-d-illustration-title-business- document-image88877337

Plantfelicity. (n.d.). Newton cradle, cause and effect concept, blue infinity Newton`s cradle 3d illustration

banner (ID 110665870) [Illustration]. Dreamstime. https://www.dreamstime.com/close-up-newton-s- cradle-newton-cradle-cause-effect-concept-blue-infinity-newton-s-cradle-d-illustration-banner- image110665870

Price, P. C., Jhangiani, R., & Chiang, I. A. (2015). Research methods in psychology.

https://opentextbc.ca/researchmethods/chapter/quasi-experimental-research/ Trochim, W. M. (2006). Internal validity. Conjoint.ly. http://www.socialresearchmethods.net/kb/intval.php

Learning Activities (Nongraded) Nongraded Learning Activities are provided to aid students in their course of study. You do not have to submit them. If you have questions, contact your instructor for further guidance and information. Review the “Interpretation Questions” and “Application Problems” at the end of Chapters 4, 5, 6, 7, and 8.

(Hafakot, n.d.)