Significance of Chronbach Alpha

To fulfill my objective of having an MBA degree, I had to do a lot of research. However impractical it may seem but my college demands 4-5 research projects to be done either as assignments or for the fulfillment of course in duration of 2 years. The quality of my research might not have been at fullest level of professionalism, partly because of research bias in data collection. But still, I have learned the most adored skill of data analysis.

I enjoy analyzing data and make meaningful conclusion out of it, but for some people it might be a daunting task to do. So, I would like to put some things in most simplistic matter to help you learn more about it. Here I am trying to explain a reliability measure, Chronbach Alpha in simple terms.

What is Chronbach’s Alpha?

Chronbach’s Alpha is one of the most widely used test to check the internal consistency of any data collected for marketing research, this is an extension to “KR-20” test which is used over dichotomous items for testing internal consistency or reliability of data collected.

What does it measure?
Chronbach’s Alpha measures the similarity of responses to individual items measuring same construct in a scale for a given case. This might be a difficult sentence to digest, so I would like to put up an example in the end to show, What it measures and how to draw conclusions from its value.

What is the significance of value of Chronbach’s Alpha?
Value of Chronbach’s Alpha is generally between 0 to 1, it might be negative sometimes. But it happens generally when the data is not re-coded properly. Thus, it is absolutely important to re-code the data measuring a particular construct and then apply the reliability test (Cronbach Alpha). Further, it should be only applied to variables measuring a particular construct. There is no use of applying it to demographics data of sample.
Values of Alpha close to 1 means there is more similar responses to various items in questionnaire. However, values close to 0 means there is very less correlation among responses of individual questions. This can be better explained through an example as shown below-

Example-
In the end, I would like to share a simple example using SPSS for data analysis, to show how Chronbach’s Alpha is calculated and interpreted. Here I first create a sample file in excel to be used for Reliability analysis in SPSS.

Let me know if this explains everything through you comments, also post some suggestions/corrections (if required).

Fast Data Entry in Excel

If you are entering lots of data from your hard copy questionnaire to Excel, then, here is a smart way I have discovered to do it, which increases your efficiency more than by 100%. I will illustrate this techniques using example where I am having 100 cases to input in different rows, each case having different response to every column (or question).

Usually you would do this by individually putting in the data  in every cell then moving across by pressing tab and enter. But here is a small trick by which you can increase your speed and avaoid pressing hundreds of tabs unnecessarily.
Trick is simple and explained in video and below-

  1. Input your data into notepad, separating each case by enter
  2. Use find and replace to create a comma separated file, ready for pasting into excel.
  3. Use Excel Text import wizard to import all data, and you are done!

Let me know through your comments about your views on this  trick, or ask me anything if it is not clear to you.