16th January 2010 at 8:57 am #5131Hassabo okashaMember
Hello everyone, I have a question & I hope to find a clear answer
My question is: I have a questionnaire contain 20 questions with Ordinal option like(1- week, 2- Pass, 3- good, 4- V. Good, 5 -Excellent)…. I did a frequency table for each variable, but the Important thing is to make a frequency table for the Overall Satisfaction, How can I do that?
Hassabo Okasha16th January 2010 at 10:31 am #5138Francisco VieiraParticipant
I think you need to review your scale. It seems that it has many positive points, for instance, good, very good and excellent, what means three in a row. Maybe it can cause a bias in the data collect and, therefore, in the results.
Francisco Vieira16th January 2010 at 12:19 pm #5137Hassabo okashaMember
Hello Francisco, and thank you for your reply, but I still confused… Suppose if I review my scale to be as you said good, very good and excellent and entered the data by coding like that (1 >>good, 2>> very good and 3 >>excellent)< as i said i have 20 variables with the same scale, it’s easy to do a frequency table for each variable sepeately but if i want to collect all values of these variable in one variable to show the overall satisfaction what are the new intervals should i use to these scales…??
Regards5th February 2010 at 8:26 pm #5135
Hi: first of all, you should process ordinal measures representing a result as median, range, and first and third quartile. All statistical programs may perform this for you!!!!! The right measure is median and range, but you should perform inference to Population, so percentile 25 and percentile 75 will give you an interval where the overall median of all of your answers lies , with 95% confidence, between it’s lower limit (the 1st quartile(percentile 25 )) and the 3rd. quartile(p.75)(upper confidence interval for the median ). You should recode in an excel sheet putting the number instead of ,i.e. “weak”). This will be the way you may have your table imported by a statistical program, and then obtain results. Good Luck!!!!, I agree also with the comment on the biais problem.6th February 2010 at 8:25 pm #5134
Look, I’m working now, and wont be available. Once you obtained your results, i.e. overall result: 3(Lower Limit 95% confidence Interval 2, Upper Limit 95%Confidence Interval 4), you must refer you result inferred to aPopulation similar to sample data you had as : “The overall……with 95% of confidence, will be between very good(2) and whatever you meaned in code 4). Thanks. Try to searh any free statistical software!!!! This is processed in place where descriptives are !!!!13th February 2010 at 1:35 pm #5133ISMAEL YOUNIS ABUJARADMember
I have a question on how to settle my model problem. I would like to raise a question ?
in data analysis using SPSS , what shall i do with my depenent variable
My dependent variable is not continuous (it is about financial profitablitiy) i measured it using BPCI (Business Performance Composite INdex) which is the ratios of ROI plus ROS plus ROA for the last three years and divided by 3 . i got it using excel , then i pasted it on SPSS …However, my independent variable is organizational innovativeness . I measured it using a 7-likert scale. do you think i have to put the dv into a scale as well or leave it as it is ?22nd February 2010 at 10:33 am #5132K.KalyanaramanMember
Your dependent variable may be taken as continuous as you are computing the same by using ROI,ROAand ROS. All you have to do is to verify whether linearity holds using a scatter diagram and use any software for getting regression, if linearity can be accepted. Otherwise suitable form of function to be chosen. You could try curve expert a free downloadable software.
- The forum ‘Default Forum’ is closed to new topics and replies.