It may be argued that the differences between statistical methods and phenomenological methods are not on a continuum between extreme polls. They may be combined at various levels.

STRENGTH OF STATISTICAL METHODS

Statistical methods are useful for a variety of purposes. They are used for constructing scales, designing studies, describing data, exploring patterns, revealing dimensionality, and predicting social phenomena. They may also be combined with qualitative methodology. Traditionally, sociologists have always tried to be in line with the principles of natural sciences and have favoured exact and standardized scientific methodology. They have considered quantitative methods as the only statistically valid and reliable methods of research. Using a deductive approach, quantitative research yields statistically reliable results. The process of quantitative research is best suited for testing hypotheses. The obtained results are utilized for establishing causal relationships and making predictions.

Quantitative research methods can be broadly classified under the four major categories:

• Descriptive or survey research,

• Correlational and causal research,

• Experimental research.

The above methods are discussed at length in the next lecture. Here it may be said that the positivists consider qualitative data as unreliable, ambiguous and doubtful, and despite criticism and attack, the quantitative approach dominates social sciences. As late as in 2009, Cole and Gramajo have used a statistical model to explain the causes of homicide rates in different cultures. Although not acknowledged by them, they have precisely used the same methodology as used by Durkheim in explaining suicide rate. The only difference is their use of logistic regression which was not known to Durkheim.