Quantitative Research Methods in Educational Research
Quantitative Approach
Quantitative research is dedicated to collecting, processing, and analyzing numerical data on certain variables to produce explanatory theories. These theories aim to establish the relationship between variables in terms of cause and effect. The hypothesis is formulated at the beginning of the investigation, and the sample selection uses mathematical or probabilistic methods.
Features
- Variability: Observations on the same issue can have different values.
- Validity: The extent to which scientific explanations of events coincide with reality. There are two forms of validity:
- Internal Validity: Ensures that external variables are taken into account or controlled.
- External Validity: Ensures that results and generalizations can be applied to other people and environments.
- Reliability: The ability of the research design to capture the true relationship between variables.
- Simplicity: The research design and analysis should not be overly complicated.
- Level of Significance: The researcher must operationalize the variables applied and determine the statistical significance of the findings.
Methods
Experimental
Experimental methods offer a high degree of control. The researcher manipulates the phenomenon under investigation and determines the values of the independent variable. This allows for full control of extraneous variables.
Quasi-experimental
Quasi-experimental methods offer a moderate degree of control. While the researcher manipulates the phenomenon, they may not have complete control over extraneous variables.
Non-experimental
Non-experimental methods offer a low degree of control. The researcher does not manipulate the phenomenon, and the relationship between variables has already occurred. The researcher can only observe and record measurements.
Ways to Define a Variable
Operationalized
Operationalization is crucial for measuring variables. There are four measurement scales:
- Nominal: Classifies individuals into categories.
- Ordinal: Operationalizes variables by indicating order or rank. Allows for the calculation of the median and ordinal correlation.
- Interval: Assigns numerical values to individuals to quantify differences. Allows for the calculation of the mean and standard deviation.
- Ratio: Similar to interval scales but includes an absolute zero, representing a total lack of the measured quality.
Constitutively
Describes the essence of an object or phenomenon, similar to a dictionary definition. These variables are not directly observable and require indicators (observable variables) for measurement and hypothesis testing.
Control of Variables
Types
- Independent Variable: The variable that is manipulated by the researcher. There are two ways to control the independent variable:
- Manipulation: The researcher systematically varies the independent variable to observe its effect on the dependent variable.
- Selection: Used when manipulation is not possible, the researcher selects participants based on their existing levels of the independent variable.
- Dependent Variable: The variable that is measured to observe the effect of the independent variable. The number of measurement points and timing can vary (pretest, posttest, or differential test).
- Extraneous Variable: Variables that are not of primary interest but can influence the dependent variable. Researchers aim to eliminate or control the influence of extraneous variables.
Quantitative Approach Design
Pre-experimental Designs
- Post-test Only with One Group: Treatment is provided, and then the dependent variable is observed.
- Pretest-Posttest with One Group: A pretest is administered, followed by the treatment, and then a posttest. The change between pretest and posttest scores is analyzed.
- Post-test Only with Two Nonequivalent Groups: Includes a control group that does not receive the treatment. The difference between the treatment group and the control group is compared.
Experimental Designs
Commonly used in physical and biological sciences, experimental designs offer a high degree of control over variables.
Quasi-experimental Designs
- Pretest-Posttest with a Nonequivalent Control Group: Groups of subjects are already established, and the researcher compares the change between pretest and posttest scores between the treatment and control groups.
- Interrupted Time Series with One Group: Multiple observations are made before and after the treatment to assess its impact over time.
- Interrupted Time Series with Control Groups: Similar to the previous design but includes a control group for comparison.
- Single-Subject Design: Similar to interrupted time series designs but involves only one subject.
Non-experimental Designs
Often used in social and educational research, non-experimental designs examine phenomena after they have occurred. The researcher cannot manipulate or control the independent variable.
Features
- Data is collected after the occurrence of the alleged causes.
- Limited control over the independent variable.
- Flexible method for establishing hypotheses.
Types
- Descriptive Studies: Provide a preliminary understanding of a phenomenon and often serve as a starting point for more in-depth investigations. Data collection methods include classroom observations, attitude scales, questionnaires, interviews, and content analysis. Data analysis typically involves statistical techniques.
- Survey Studies: Utilize questionnaires or interviews to collect data. Survey research can be used for:
- Gaining an initial understanding of a phenomenon.
- Describing or predicting the nature of an educational phenomenon.
- Identifying norms or standards for comparison.
- Determining relationships between specific events.
Prerequisites for Survey Design
- Clearly defined research objectives.
- Identification of the target population.
- Awareness of available resources.
Phases of Survey Research
- Defining the objectives of the survey.
- Planning the survey (using concept maps or specification tables can be helpful).
- Developing and selecting questions.
- Analyzing the quality of the questions.
- Analyzing the reliability and validity of the questionnaire (pilot testing, expert review).