Session I & II- Business Analytics
In the introductory classes of
Business Analytics we were familiarized with various basics and fundamentals of
using the software package – SPSS.
Through understanding the ways of
using the SPSS software, we were also made to understand some concepts and statistical
tools and how they could be “analyzed” to gain better insight into the subject
at hand (as opposed to just processing the data).
The first fundamental explained was
the concept of variables:
The next concept was of data:
Data can be:
- Nominal: only for name sake. One has the freedom to assign any number to a variable. Eg: gender – I can make Male & Female 1&2 or 5&6 or 4&5 or anything. Eg: nos. on the back of player jerseys. Eg: cell phone nos.
- Ordinal: shows order (eg if you’re showing order of education. Can be 1234 or 4321 but needs to show order). Eg: Designation (Pres, VP, Asst…etc.)
- Scale:
Somewhere one can tell the EXACT magnitude of difference between two
observations. Time, distance.
Classification of techniques:
- 1st level analysis: frequencies, cross-tabs, OLAP cubes
2nd
level analysis: multi-variate techniques
- Univariate – use/ show only one variable at a time. Pie chart (various values of the same variable) A pie chart can be bivariate if one bring thickness of the pie into the picture.
Bivariate –
correlation, t-test
Multivariate –
bubble diagram
More the no. of
variables, more insight can be gained.
Finally we completed an exercise
using our knowledge of the above concepts and the basics of SPSS to begin
proving a hypothesis for a given set of data.
From this a few pointers that
emerged were:
a)
In cross-tabs: Null Hypothesis always same ie – “No
relation between the two variables under consideration”
b)
Check significance value in chi-square table. If that
number is less than .05 there is a significant relationship between the
variables and therefore we reject the null hypo.
By
Urvashi Singh
SIBM Bangalore
14057
No comments:
Post a Comment