ployee ’ s salary can be framed in the following ways —
• According to you , how is your salary ? Good / Bad ( Nominal scale )
• Please indicate your level of salary . Very High / High / Moderate / Low . ( Ordinal Scale )
• Indicate your salary level on a scale of 1-10 . ( Interval Scale )
Very Low Very High 1 10
• What is your gross salary in Indian rupee per month ? ( Ratio Scale )
Before elaborating on why so much importance is given on framing any question in different type of scale , let ’ s have some idea about the various types of data analysis .
Most basic type of statistical analysis is “ Descriptive analysis ”, which is used to summarise the responses from large number of respondents in a few simple statistical measures . Data can be described in terms of “ Frequency ”, “ Central Tendency ’ and “ Statistical Distribution ” followed by a set of data . Mean , median , mode , range , standard deviation , variance are widely used in descriptive statistics . Data description can be done with the help of tabulation , graphs , charts and cross tabulation . Mean , median and mode gives an idea about the central tendencies of the data and range and standard deviation depict the variability of data . Following table gives an idea about various measures of descriptive analysis to be used for various types of scales .
Once the data description is done inferential statistics plays a role to test hypothesis . It means some inferences can be drawn about some population based on the observation of a small sample . For example a research on work life balance of Indian working women does not need the entire Indian working women to be studied , it can be done by taking a sample consisting of small number of females who are working in various sectors in a geographical area . Hypothesis testing is utilized to do the inferences . Hypothesis is a formal statement of explanation stated in a testable form . Hypotheses are tested to draw a conclusion about population or about dependency of variables . Depending on the number of variables to be studied , there types of statistical analysis are there —
• Univariate statistical analysis- hypothesis involving only one variable . E . g . Working women in IT sector are highly stressed .
• Bivariate statistical analysis- tests hypotheses involving two variables . E . g . Flexible timing of work reduces stress level of female workers .
• Multivariate statistical analysis- tests hypotheses involving more than two variables . E . g . Flexible timing of work , number of family members , age of the employee , inter-organisational relationship are having impact on stress level of female workers .
Nominal Scale Ordinal Scale Interval Scale Ratio Scale
Mode , Frequencies
Mode , Median , Frequencies , Range
Mode , Median , Frequencies , Range , Mean , Standard Deviation
Mode , Median , Frequencies , Range , Mean , Standard Deviation , Geometric mean , Harmonic mean