## steps in correlational research

Correlational research uses numerical data to explore relationships between two or more variables. The degree of relationship is expressed in terms of a coefficient of correlation.

Correlational research describes what exists at the moment (conditions, practices, processes, structures etc.) and is therefore, classified as a type of descriptive method. Nevertheless, these conditions, practices, processes or structures described are markedly different from the way they are usually described in a survey or an observational study. Correlational research comprises of collecting data to determine whether, and to what extent, a relationship exists between two or more quantifiable variables. Correlational research uses numerical data to explore relationships between two or more variables. The degree of relationship is expressed in terms of a coefficient of correlation. If the relationship exists between variables, it implies that scores on one variable are associated with or vary with the scores on another variable.

The exploration of relationship of the relationship between variables provides insight into the nature of the variables themselves as well as an understanding of their relationships. If the relationships are substantial and consistent, they enable a researcher to make predictions about the variables.

Correlational research is aimed at determining the nature, degree and direction of relationships between variables or using these relationships to make predictions. Correlational studies typically investigate a number of variables expected to be related to a major, complex variable. Those variables which are not found to be related to this major, complex variable are omitted from further analysis. On the other hand, those variables which are found to be related to this major, complex variable are further analysed in a causal-comparative or experimental study so as to determine the exact nature of the relationship between them.

In a correlational study, hypotheses or research questions are stated at the beginning of the study. The null hypotheses are often used in a correlational study. Correlational study does not specify cause-and-effect relationships between variables under consideration. It merely specifies concomitant variations in the scores on the variables. For example, there is a strong relationship between students’ scores on academic achievement in Mathematics and their scores on academic achievement in Science. This does not suggest that one of these variables is the cause and the other is the effect. In fact, a third variable, viz., students’ intelligence could be the cause of students’ academic achievement in both, Mathematics and Science.

Correlational study is designed (a) to determine whether and how a set of variables are related, or (b) to test the hypothesis of expected relationship between among the set of two or more variables. The variables to be included in the study need to be selected on the basis of a sound theory or prior research or observation and experience. There has to be some logical connection between the variables so as to make interpretations of the findings of the study more meaningful, valid and scientific. A correlational study is not done just to find out what exists: it is done for the ultimate purpose of explanation and prediction of phenomena. If a correlational study is done just to find out what exists, it is usually known as a ‘shot gun’ approach and the findings of such a study are very difficult to interpret.

The minimum acceptable sample size should be 30, as statistically, it is regarded as a large sample. The sample is generally selected using one of the acceptable sampling methods. If the validity and the reliability of the variables to be studied are low, the measurement error is likely to be high and hence the sample size should be large. Thus it is necessary to ensure that valid and reliable tools are used for the purpose of collecting the data. Moreover, suppose you are studying the relationship between classroom environment and academic achievement of students. If your tool measuring classroom environment focuses only on the physical aspects of the classroom and not its psycho-social aspects, then your findings would indicate a relationship only between academic achievement of students and the physical aspects of the classroom environment and not the entire classroom environment since the physical aspects of the classroom environment is not the only comprehensive and reliable measure of classroom environment. Thus the measurement instruments should be valid and reliable.

The basic design of a correlational study is simple. It requires scores obtained on two or more variables from each unit of the sample and the correlation coefficient between the paired scores is computed which indicates the degree and direction of the relationship between variables.

In a study designed to explore or test hypothesized relationships, a correlation coefficient is interpreted in terms of its statistical significance.

Correlational research is of the following two types:

(a) Relationship Studies: These attempt to gain insight into variables that are related to complex variables such as academic performance, self-concept, stress, achievement motivation or creativity.

(b) Prediction Studies: These are conducted to facilitate decisions about individuals or to aid in various types of selection. They are also conducted to determine predictive validity of measuring tools as well as to test variables hypothesized to be predictors of a criterion variable.

The exploration of relationship of the relationship between variables provides insight into the nature of the variables themselves as well as an understanding of their relationships. If the relationships are substantial and consistent, they enable a researcher to make predictions about the variables.

Correlational research is aimed at determining the nature, degree and direction of relationships between variables or using these relationships to make predictions. Correlational studies typically investigate a number of variables expected to be related to a major, complex variable. Those variables which are not found to be related to this major, complex variable are omitted from further analysis. On the other hand, those variables which are found to be related to this major, complex variable are further analysed in a causal-comparative or experimental study so as to determine the exact nature of the relationship between them.

In a correlational study, hypotheses or research questions are stated at the beginning of the study. The null hypotheses are often used in a correlational study. Correlational study does not specify cause-and-effect relationships between variables under consideration. It merely specifies concomitant variations in the scores on the variables. For example, there is a strong relationship between students’ scores on academic achievement in Mathematics and their scores on academic achievement in Science. This does not suggest that one of these variables is the cause and the other is the effect. In fact, a third variable, viz., students’ intelligence could be the cause of students’ academic achievement in both, Mathematics and Science.

**Steps of a Correlational Research****1. Selection of a Problem:**Correlational study is designed (a) to determine whether and how a set of variables are related, or (b) to test the hypothesis of expected relationship between among the set of two or more variables. The variables to be included in the study need to be selected on the basis of a sound theory or prior research or observation and experience. There has to be some logical connection between the variables so as to make interpretations of the findings of the study more meaningful, valid and scientific. A correlational study is not done just to find out what exists: it is done for the ultimate purpose of explanation and prediction of phenomena. If a correlational study is done just to find out what exists, it is usually known as a ‘shot gun’ approach and the findings of such a study are very difficult to interpret.

**2. Selection of the Sample and the Tools:**The minimum acceptable sample size should be 30, as statistically, it is regarded as a large sample. The sample is generally selected using one of the acceptable sampling methods. If the validity and the reliability of the variables to be studied are low, the measurement error is likely to be high and hence the sample size should be large. Thus it is necessary to ensure that valid and reliable tools are used for the purpose of collecting the data. Moreover, suppose you are studying the relationship between classroom environment and academic achievement of students. If your tool measuring classroom environment focuses only on the physical aspects of the classroom and not its psycho-social aspects, then your findings would indicate a relationship only between academic achievement of students and the physical aspects of the classroom environment and not the entire classroom environment since the physical aspects of the classroom environment is not the only comprehensive and reliable measure of classroom environment. Thus the measurement instruments should be valid and reliable.

**3. Design and Procedure:**The basic design of a correlational study is simple. It requires scores obtained on two or more variables from each unit of the sample and the correlation coefficient between the paired scores is computed which indicates the degree and direction of the relationship between variables.

**4. Interpretation of the Findings:**In a study designed to explore or test hypothesized relationships, a correlation coefficient is interpreted in terms of its statistical significance.

Correlational research is of the following two types:

(a) Relationship Studies: These attempt to gain insight into variables that are related to complex variables such as academic performance, self-concept, stress, achievement motivation or creativity.

(b) Prediction Studies: These are conducted to facilitate decisions about individuals or to aid in various types of selection. They are also conducted to determine predictive validity of measuring tools as well as to test variables hypothesized to be predictors of a criterion variable.