How To Analyse An Open Ended Questionnaire - XLS Download

I am doing my summer project on ANLYSIS OF THE EXIT QUESTIONNAIRES OF EX – EXPLOYEES OF MY COMPANY WHO LEFT THE COMPANY IN THE LAST FINANCIAL YEAR

Now the problem is that it is totally an open ended questionnaire that I need to analyze. There are about 72 employees –of different departments - who left last year. I have typed some samples questions below. Can anyone guide me as to how to analyze it…

Is it fine if I can convert the open ended questions into closed ended ones by making assumptions and then do a normal analysis ….I mean would that be a wrong way of analyzing ….

For example For question 5 below can I put a 5 point scale highly satisfied, satisfied, neutral, dissatisfied , highly dissatisfied- and then analyse it in an excel sheet

I would appreciate if anyone could help me out with this asap coz only after that I can start my project



Employee name. ________________________

1.Reasons for separation (this is an objective type question. So it’s fine. )

Ø Better compensation

Ø Superiors

Ø Less to travel to work

Ø Working environment

Ø Higher education

Ø Any other



2How did you feel about following

Ø Career progress with the company: __________________________________________________ __________________________________________________ ______________

Ø Training and development inputs received

__________________________________________________ __________________________________________________ ______________

Ø HR Policies and systems

__________________________________________________ __________________________________________________ ______________

Ø Compensation policies

__________________________________________________ __________________________________________________ ______________

3. Did you feel following was satisfactorily explained?

Company policies: __________________________________________

Rules and regulations: _______________________________________

4. Did he like the work he was doing?

__________________________________________________ _______

(Sir, Here some have said just yes or no… some have elaborated ….so do I put a yes/no type of question for analysis or do I put a 5 point scale like highly satisfied, satisfied, etc) __________________________________________________ _______

5. How did you feel about the communication within the company?

__________________________________________________ __________________________________________________ ______________

6. Did you get along well with the superiors? How did the superiors handle grievances?

__________________________________________________ __________________________________________________ ______________

7 any suggestions to improve:

__________________________________________________ __________________________________________________ ______________

__________________________________________________ __________________________________________________ ______________

__________________________________________________ __________________________________________________ ______________

You can try to categorise the comments for every question and say you have 5 categories for Q2 then use usual stats to analyse eg 5 respondents for category 1, 20 respondents for category 2 etc.
Open ended is definitely more work but richer info gathered.
Just a suggestion.

thanx for the suggestion
so are you saying that its right to convert open ended to closed ended ques

you need to go through the answers and then try to define it cause i might be a sort of person who is easily satisfied while you may neede more inputs to stay satisfied...instead use the open ended answers to find key words and then do the analysis accordingly.

Well, actually its trying to make the open ended responses into statistics for analysis. But once you've analysed - say Q3's highest dissatisfaction is in area of category 2, you can dig back into the qualitative info to undersand what's the problem.
Don't know if its valid but its what we do! Good luck!

:D Hi Tania.

The 'Scale' you are using is a 'Likert' Rating system : e.g. Highly Satisfied, Satisfied etc .. to Highly Dissatisfied which can be further scaled. I am just sharing my experience and stating the relevant portions which you can correlate .. though my perspective was on 'Suggesting a better system' rather than analysis, I think you may proceed as under ..

:idea: As per Likert Rating Scale 5 possible answers (Responses) are there: Highly Satisfied, Satisfied, Neutral, Dissatisfied, Highly Dissatisfied., this you can change into 3, 4 or 7 point rating scale. To do this (I did it), assign Scores to 5 point responses, e.g. 0 for Highly Dissatisfied and similarly .. 4 for Highly Satisfied .. its useful to have a quantitative measure to switch or to further scale another rating system. Once you do it, use Attriburte Measure : Total Score obtained by all Ex-employees / Total possible Score .. to keep it simple - higher the score, better the rating. Now, if you have marginal scores like 0.57 or say 0.48 or 0.6, just compliment the analysis with your qualitative data. As these are Human Perceptions, strict quality control or cross sectional analysis is not possible. If you are switching from 5 point to 2 point scale, it would be easier. Vice versa you will have to depend more on qualitative data to translate 2 point scale to 5 point scale ..

:arrow: That is to say .. I think .. if you concentrate on End Product of that Analysis I hope you will be able to find a good solution. If you are unsure, make these seperate lists: What is to be done/ improved, What Data is required, From which data analysis is to be made .. and you can identify certain Areas in which you can take individual Attribute scores, e.g. Quality of Work Life, Communication, Decision Taking and Leadership, Internal Customer Focus, Compensation & benefits, Ethical values etc .. this will give a more comprehensive picture of exit scenario and may suggest certain areas in which improvement are necessary to retain talents !

:idea: My work was a bit different from yours in which Exit Process improvement was a small area .. also it was for Managemet Cadre in an Iron & Steel Industry .. I hope this work to be of some use !

8) You can do a synthesis of various alternate methods, take the best elements and prepare your own reasoning system ..

Wishing you a very best of luck .. :D

Regadrs,

thesysthink

thanx all u people for the suggesstions .
i have divided the entire lot of 75 questionnaires into 7 departments.
now i analysed 1 departmant of abt 20 ex-employees and was successful in finding out a common factor for attrition in that department .
now the problem is that in another dept consisting of only 16 people who left last year ,wen i went to analyse , each person has a different reason for leaving the org. for eg some hv personal probs, some carer , some compensation , etc. due to this i would be able to say that attrition here is due to a specific prob.
so it is better that i dont do department wise at all and do it a one org as a whole .

try out using dummy tables , that is use the excel sheet in figuring out and cross tabulating 2 factors . for example age of the employee in x axis and another relavent factor in the y axis , do it for many factors which r esssential to be analysed
raaj

thanx thesysthink .
that was real sood insight thankyou very much.
have you heard of SPSS software . iit a software used for analysis. if anyone can forward a linf to downoad spss to [Login to view] it will be great
thanx

:D Dear Tania,

Unfortunately SPSS cannot be legally downloaded from ANY website (you need to purchase a license) ..

What I suggest is use the 'Data Analysis' Solver packs in Microsoft Excel. Hopefully you won't require complex multivariate models, most Parametric stats can be tackled by Excel .. it needs to be installed seperately if you cannot see it in Excel now ...

The Solver is usually found in the same Installation CD from which the MS Office is installed .. just discuss it with your system admin .. if you are using version 2000 or higher it should be no prob for Installation.

Alternately there is MSTAT Package .. only prob is that its a very old DOS based application that reads data from Floppy Drive directly .. this is a pretty common utility which is handy to use. I have used SPSS 2 years ago .. lost touch a bit. There are Statistica / STATCAT, but in my opinion they are to be used only if you are planing to use Mathematical prog./ Opsearch tech.'s in your project ..

Best of luck ! :D

Regards,

thesysthink