What are Regression Analyses?

Regression Analyses, as explained by DJS Research Ltd.

Regression analyses are statistical tools used, in their most basic form, to detect links between two or more phenomena. An example from a market research perspective would be if one were seeking to find out whether those students who spent longer doing their History homework received better grades in that subject. This data can be plotted as points on a graph, where the x-axis is the average number of hours per week a student studies, and the y-axis represents exam scores out of 10. These points will usually be scattered on the graph; the regression analysis generates the line that best summarises the distribution of points. The typical distance between the line and all the points (the “standard error”) indicates whether or not the regression analysis has captured a strong relationship: the closer the line to the points generally, the stronger the relationship. 

However, even a line that sits very closely to all the points may not be proof of a definitive link between the measured variables. Perhaps those history students who did well may have more aptitude for analytic writing, or perhaps they have a deep rooted interest in the subject which encourages them to study but without this extra study being of any real relevance to the exam. These complexities are increased as the number of variables increases. Still, regression analysis – at the very least – points to things which should to be looked at in further detail.

For further information please visit DJS Research, a full service market research agency. Alternatively, feel free to call them on +44 (0) 1663 767857

Follow DJS Research Ltd: On Twitter :: On Facebook :: On Linkedin :: On You Tube