Author: Dr. Prasamita Mohanty

[shc_shortcode class=”shc_mybox”]

Director, Centre for the Study of Social Exclusion and Inclusive Policy, BBAU, Lucknow.

Published in: Contemporary Researches in Education, Edited by Dr.Asha J.V. and Naseerali M.K.



In the field of education, neither the treatment nor the ability to manipulate the condition is conducive to an experiment. Thus researcher prefers to conduct correlational research to study the relationship between two or more variables or to predict an outcome. Correlational Research concerned with determining the extent of relationship existing between the variables. The magnitude of relationship is determined through the use of co-efficient of correlation. In correlation Research, the investigators use the correlation statistical test to describe and measure the degree of relationship between two or more variables or sets of scores. In this research, the researcher do not attempt to control or manipulate the variables as an experiment, instead they relate and using the correlation statistics.

When do we use Correlational Research?

When two or more variables influence each other such as relationship between Academic Achievement and Adjustment. This type of research predict an outcome such as inter student relationship, teacher- student relationship, classroom adjustment and academic achievement influence student achievement. Researcher can apply statistical technique by calculating Correlation test.

How did Correlational Research Develop?

The origin of correlational research developed by Statistician in the late 19th Century(Cowles,1989). Although British Biometrician articulated the basic ideas of “co-relation” during the last half of 1800s.Pearson used illustrations from Darwin’s Theory of Evolution and Sir Francis Galton’s ideas on heredity for experiment the concept of correlation. In1897, Yule (Pearson’s student) developed solutions for correlating two, three and four variables. In 1904, Spearman published ideas about a correlation matrix to display the   coefficients. Fisher (1935) pioneered significance testing and ANOVA, important statistical ideas for studying the difference between observed and predicted scores in correlational analysis. It was not until1963 that Campbell and Stanley provided new impetus to correlational research. In using correlational research, they encouraged researchers to both recognize and specify the extensive threats to validity inherent in this form of research.

During 1970 and 1980 with the advent of computers, improved knowledge about measurement scales and the need to study complex associations among many variables, quantitative researchers initiated correlational studies.Instead of the physical control available to experimental researchers through techniques such as randomization and matching, correlational researchers sought control through statistical procedures. With computers, they could statistically remove the effect of a large no. of variables to examine the relationship  among a small set of variables. They could explore the combination of variables(e.g.,age,gender and SAT Scores) and an outcome. From simple regression- the analysis of the variability of a single dependent variable by a single independent variable- the technique of using multicle regression to analyze the collective and separate effects of two or more independent variables on a dependent variable emerged.


Correlation Matrix:

The correlation study is relatively easy to design and conduct. It collects two or more sets of measurement and prepare correlation matrix. A correlation matrix presents a visual display of the correlation coefficient for all variables in a study. In this display, all variables listed in both horizontal and vertical column in the table. When interpreting the score it is important to identify the direction whether move in same or opposite direction. The direction of relationship may be positive or negative; the degree of relationship may be vary from perfect to high, to average, to no relationship; the relationship may be linear or curvilinear. When low score of one variable relate to the low score on a second variable then positive linear relationship exists. When low score in one variable relate to high scores on the other variable then negative linear relationship exists. When a particular score of one variable does not predict any information about the possible score on other variable then nonlinear relationship exists.

Degree of Correlation:

The degree of correlation means the association between two variables or sets of scores is a correlation coefficient of -1.00 to +1.00 indicating no linear association. This association between two sets of scores reflects whether there is consistent, predictable association between the scores. According to Cohen and Manion (1994) the table of correlation coefficient interprets:

  • .20-.35: When correlations range from .20 to .35, there is only a slight relationship. This relationship may be slightly statistical significant but little value in prediction studies.
  • .35-.65: When correlation are above .3 they are useful for limited prediction
  • .66-.85: When correlation falls into this category, good prediction can result from one variable to the other. Coefficient in this range would be considered very good.
  • .86 and above: Correlations in this range are typically achieved for studies of construct validity or test -retest


A number of correlational techniques are employed for various types of data.

Correlation coefficient Type of scale
1. Pearson product moment The characteristic of both variables are expressed in interval scale
2. Spearman rank difference The characteristic of both variables are expressed in  ordinal scale
3.Point Biserial One variable on interval scale, the other a genuine dichotomous variable on a nominal scale
4.Biserial One variable on interval scale, the other an artificial dichotomy
5.Tetrachoric The characteristic of both variables are expressed in nominal scale, artificial dichotomy used with both variables and both the variables have continuous distributions
6.Phi-coefficient The characteristic of both variables are expressed in nominal scale and genuine dichotomy used with both variables.
7.Contingency Coefficient Both the  variables are  classified into two or more categories and the characteristic of  both variables  are expressed in nominal scale


  1. Determine the Research Problem

A correlational study is used when a need exists to study a problem requiring the identification of the degree of association between two sets of scores. The problem which explain the complex relationship of multiple factors, identify the type of association and predict the outcome.

Some sample research questions in correlational study are:

  • Is Creativity is related to Intelligence of school children?(association of two variables)
  • What factors are responsible for higher student learning?(complex relationship)
  • Does class test score predict the final result of the annual examination?(prediction)
  1. Identify the sample

The next step of conducting the correlational study is to identify the sample and the sample size. The larger size contributes to less error variance and better representativeness.

  1. Identify two or more variables

The basic idea of correlational research is to compare the participants in two or more variables in order to prove the reliability and validity.

  1. Collect Data

The next step is to administer the tools and collect at least two sets of data from each individual. This will predict single outcome from a single predictor variable and multiple predictors from complex relationship. Other factors that might affect the study are lack of standard administration procedures, condition of testing situation and expectation of participants.

  1. Analyse the Data

The objective of correlational research is to describe the degree of association between two or more variables. The researcher uses statistical procedures to determine the strength of relationship as well as its direction. The analysis begins with coding data and transfer to a computer file. Then the researcher needs to determine the appropriate statistics techniques to apply and calculate the result.

  1. Interpret the Result

The final step in conducting the correlational study is interpreting the meaning of result. This requires the magnitude and direction of the result in a correlational study, considering the impact of intervening variables in a partial correlation study, interpreting the regression weights of variables in a regression analysis and developing a predictive equation for predictive study.

How to Evaluate Correlational Study?

  • An adequate sample size for hypothesis testing
  • The display of correlation results in a matrix.
  • An interpretation about the direction and magnitude of the relationship
  • The choice of an appropriate statistics for analysis
  • The identification of predictor and the criterion variable


Best, John W.,& James, V. Kahn. (2008). Research in Education, New Delhi, Prentice Hall of India Pvt. Ltd.

Good, Carter V. (1966) Essentials of Educational Research: Methodology and Design. New York,Appleton-Century Crofts.

John, W. Creswell. (2012). Educational Research, New Delhi, Pearson Education Inc.

Kerlinger, Fred N. (2008). Foundations of Behavioural Research, New Delhi, Surjeet Publication

Koul, Lokesh. (2013). Methodology of Educational Research, Noida, Vikas Publishing House Pvt.Ltd

Sharma, A.K. (1997). Educational Research in India: An Overview in fifth Survey of Educational Research;1988-92, New Delhi,NCERT.



Leave a Reply

Your email address will not be published. Required fields are marked *