Table of Contents
What is Research?
A research is a systematic investigation that uses disciplined methods to answer questions and solve problems. The main goal of research is to develop, refine, and expand knowledge.
A research can also be defined as a systematic process of collecting, analyzing, and interpreting information (data) in order to increase our understanding of a phenomenon about which we are interested or concerned.
According to Y.K. Singh (2006), a research is not just collecting and tabulating facts, it must have the following general characteristics:
It gathers new knowledge or data from primary or first-hand sources.
It places emphasis upon the discovery of general principles.
It is an exact systematic and accurate investigation.
It uses certain valid data gathering devices.
It is logical and objective.
The researcher resists the temptation to seek only the data that support his hypotheses.
The researcher eliminates personal feelings and preferences.
It endeavors to organize data in quantitative terms.
Research is patient and unhurried activity.
The researcher is willing to follow his procedures to the conclusions.
Research is carefully recorded and reported.
Conclusions and generalizations are arrived at carefully and cautiously.
Quantitative vs. Qualitative Research
The data collected in research studies take one of two general forms; quantitative and qualitative research. Qualitative research is an investigation whose data do not consist of numerical information, but rather of narrative or textual information (interviews, focus groups, documents, etc.). In general, qualitative research seeks to understand a phenomenon within a real-world context through the use of non-numerical data and inductive reasoning.
On the other hand, Quantitative research involves the process of objectively collecting and analyzing numerical data to describe, predict, or control variables of interest. Quantitative research seeks to understand the influential, causal or correlational relationship between variables through testing hypotheses to support or reject it based on statistical analysis and deductive reasoning.
What is Experimental Research?
Quantitative researches can be mainly classified into to categories; Experimental researches (controlled trials) and Nonexperimental researches (observational studies).
An experiment is defined as a study in which purposeful changes are made to subjects or experimental units under controlled conditions in order to identify changes in a response.
In experimental researches, we apply treatments to experimental units and then measure one or more responses of these units. While, an observational study simply collects data under varying conditions without intervening or imposing any changes (e.g. a cross-sectional survey).
Concepts You Should Understands
In experimental researches, there are some of the important terms and concepts we should understand. For instance, "randomization" which is a concept in sampling methods and research design; this concept says that each individual of the population under study should has the equal chance or probability to be selected in the sample selection process to constitute the sample. Also, randomization includes the use of a known probabilistic mechanism for the assignment of treatments to experimental units.
Sample selection is a collection of techniques that experimental researchers can use to select the individuals or subjects to conduct the study on.
Another important term is "treatments which represent the levels of a factor that we want to study. For example, if an experiment compared the drug dosages 50, 100, and 150 mg, then the factor "drug dosage" would have three levels (treatments): 50 mg, 100 mg, and 150 mg that was applied on the experimental units. In statistical analysis context, this type of factors (e.g. drug dosage) is called the independent variable.
Experimental units are the subjects or things to which we apply the treatments on. These could be plots of land receiving fertilizer, groups of students receiving different educational techniques, or batches of feedstock processing at different temperatures.
Replication (replicates) are a group of experimental units that receives the same treatment; and each unit of this group is called "replicate". Replication is essential for estimating experimental error and the greater the number of replications, the greater is the precision in the experiment.
Responses are outcomes that we observe after applying a treatment to an experimental unit. Simply, the response is what we measure to judge what happened in the experiment; we often have more than one response. For example, a response could be a nitrogen content of corn plants treated with NPK fertilizer. A response could be the children's IQ scores after applying a particular preschool program. In statistical analysis context, the response (e.g. nitrogen content, Children's IQ score) is called the "dependent variable" because its status depends to some degree on the status of the independent variable (e.g. NPK fertilizer, preschool program).
Steps in Planning an Experimental Research
In experimental studies, the research involves the following sub plans or steps:
1. Define the research problem, questions and objectives. Experimental researchers begin by identifying an interesting and/or significant research problem and formulating good research questions and clearly defining the objectives.
2. Formulating the hypothesis. Some tentative solutions are given for the problem when these solutions are based on certain rationale they are termed as hypothesis. Simply, hypotheses are reasonable predictions of expected outcomes.
3. Identifying the variables under investigation. In this step you should clearly and precisely indicate the type of variables under investigation (e.g. independent and dependent variables) and how you intend to measure dependent variables or determine their values.
4. Identifying the population of interest and the sampling plan. The sampling plan specifies in advance what characteristics the study participants should possess, how the sample will be selected from the population and how many subjects there will be.
5. Selecting a research design and developing protocols for the intervention. This includes selecting the experimental treatments (independent variables) and assigning them to experimental units or study participants.
6. Collecting the data and preparing them for analysis. This includes collecting the data, arranging them in tabulated format and cleaning the data. fixing or removing corrupted, incorrectly formatted, duplicate, and incomplete data, as well as dealing with outliers.
7. Analyzing the data and interpreting the results. Researchers need to analyze their data to get the results that answers the research questions. Quantitative data is analyzed through statistics (i.e. statistical analyses) to obtain results to visualize them in tables, charts, scatter plot, etc., and that is used to answer research questions, test hypotheses and consequently draw conclusions. The most statistical analysis used in experimental researches are descriptive statistics (e.g. the mean, median, standard deviation) and inferential statistics; such as Student t-test, Analysis of Variance (ANOVA), Correlation, Regression analysis, Chi square test, etc.
What is Experimental Design?
Planning and designing an experiment properly is very important in order to ensure that the right type of data and a sufficient sample size and power are available to meet the research objectives or answer the research questions of interest as clearly and efficiently as possible and consequently to obtain strong conclusions.
According to Polit and Beck (2010), the research design is the overall plan for obtaining answers to the questions being studied and for handling various challenges to the worth of the study evidence.
In experimental studies, the research design can be defined as a methodology in which a researcher controls variables (factors) systematically. Researchers may alter the level, intensity, frequency, or duration of a variable to examine any resulting change in response or behavior in the subject or participant being studied.
There are many experimental designs the researcher can select and use to conduct an experimental research. The following are common research designs used in experimental studies:
Pretest-posttest design.
Posttest only design.
Quasi-experimental design.
Completely randomized design (CRD).
Randomized complete block design (RCBD).
Factorial design.
Experimental Design Examples
We previously defined the experimental design as a methodology in which a researcher controls variables (factors) systematically. Simply, an experimental design refers to how the treatments are assigned to the experimental units to test causal relationships—to test whether the intervention caused changes in the dependent variables under study. To get deeper understanding on what the experimental designs and how they use the randomization concept, let's explore the following two designs.
1- Pretest–Posttest Control-Group Design
In pretest–posttest control-group design, experimental units of study (e.g., patients with certain disease, or a particular plant or animal species) are randomly assigned to either an experimental group or a control group. The experimental group is observed (pretest) and subjected to the experimental treatment, and observed again (posttest).
Pretest are the measurements taken on the experimental units before they are involved in some treatment, while posttest are the measurements (responses) taken after the treatment.
The basic format for the pretest–posttest control-group design is as follows:
Group 1 is the experimental group that will receive the experimental treatment. For example, a treatment could be giving a certain medication to patients or a certain dose of growth stimulant to a plant species under study.
Group 2 is the control group that receives no treatment or may receive a treatment that has no active properties, such as placebo (e.g. a sugar pill) because the control group should isolated from any influences of the experimental treatment.
This design enables us to do two main comparisons; we can statistically compare between pretest and posttest data of the experimental group (Group 1) to know, for example, whether the treatment has an effect or not, and we can make another comparison between experimental group (Group 1) data and the control group (Group 2) data.
2- Completely Randomized Design
Completely Randomized Design (CRD) is the simplest randomized experiment for comparing several treatments. This design is very important because many of the concepts and methods learned in the CRD context can be transferred with little or no modification to more complicated designs.
The CRD is a type of experimental design where the treatments (levels of a factor) are assigned completely at random so that each experimental unit receives only one treatment.
Let's assume that we have 3 treatments called "A, B, C" and we have 12 uniform experimental units. In CRD, we simply assign each treatment randomly to 4 experimental units (4 x 3 = 12) as shown in following layout.
Note that each treatment is randomly applied on 4 experimental units (replicates). In CRD design, we need at least 3 replicates for each treatment.
The CRD design can be used when the experimental units are believed to be homogeneous (uniform). Because of the homogeneity requirement, it may be difficult to use this design for field experiments.
Summary
Planning and designing an experiment properly is very important in order to ensure that the right type of data and a sufficient sample size and power are available to meet the research objectives or answer the research questions of interest.
Experimental research is a kind of study that rigidly follows a scientific research design (experimental design). The experimental design is a methodology in which a researcher controls variables (factors) systematically to test causal relationships—to test whether the intervention caused changes in the dependent variable
There are many experimental designs used in experimental research including posttest only design, pretest-posttest design, Quasi-experimental design, CRD and CRBD design.
References
Beins, B. C., & McCarthy, M. A. (2012). Research methods and statistics. USA: Pearson Education.
Leedy, P. D., & Ormrod, J. E. (2015). Practical research: Planning and design. USA: Pearson Education.
Mendenhall, W. M., & Sincich, T. L. (2016). Statistics for Engineering and the Sciences Student Solutions Manual (6th ed.). USA: Taylor & Francis Group, LLC.
Oehlert, G. W. (2010). A first course in design and analysis of experiments. University of Minnesota, USA.
Polit, D. F., & Beck, C. T. (2010). Essentials of nursing research: Appraising evidence for nursing practice (7th ed.): Lippincott Williams & Wilkins.
Singh, Y. K. (2006). Fundamental of research methodology and statistics. New Delhi, India: New Age International.
Comments