Sufficient Sample Size to represent the population of the study???????!!!!!!!!

Replies to This Discussion

Permalink Reply by Abiodun .A. Joachim on February 3, 2010 at 10:57am

Henry Mintzberg did a classical work on Managerial Roles. have you ever wondered what his sample size was? Analyse the geographical spread of your respondents so far, do more of follow up so that you could reasonably reflect good follow up effort on the respondents I am sure some would reply afterward you can proceed with your analysis? You could also search the literatures as to the response rate in similar works and in particular in the locations you are considering now

▶ Reply to This

Permalink Reply by ISMAEL YOUNIS ABUJARAD on February 4, 2010 at 4:48pm

Thank You so much Sir for the information you've provided. Well, could you share with me Henry Mintzberg work and the sample size he managed to get. Well, I have done my best in following up with them, but they simply do not want to respond for certain reasons. The response rate in the housing industry in Malaysia is terribly low... Can I get a background of what you are doing and where you are from.Thanks alot

▶ Reply to This

Permalink Reply by Abiodun .A. Joachim on February 8, 2010 at 9:29am

Ismael,

Mintzberg studied only 5 managers. The paper written in the 70s is titled Managerial Works: Analysis from Observation . If you visit www.jstor.org you might be able to download the paper from the jounal Management Science 18(2) 1971 Press the searh button and type the article title or the Authors name.Searh for earlier works in the area you are working now within and outside Malaysia so as to compare the response rate with yours. I believe you might be able to find a basis to justify what you are doing. If I may ask how many months have you been on the work? You still need to stretch yourself in follow up. Personal visit or use insiders within those organisations that are sympathetic to your work and efforts. I currently teach in a Univesity in West Africa.My research focus in in Health Care Management. Quantitative Analysis, Operatios and Strategic Mangement

▶ Reply to This

Permalink Reply by Abiodun .A. Joachim on February 8, 2010 at 10:20am

Ismael, Pls check out the discussion forum initiated by David Kleist on Sept 3,2009 It does appear some of the matters discussed are relevan to you

▶ Reply to This

Permalink Reply by Muhammad Zubir on February 12, 2010 at 11:16pm

It smell quatitative......

▶ Reply to This

Permalink Reply by ISMAEL YOUNIS ABUJARAD on February 13, 2010 at 1:33pm

Dear Mr. Muhamed Zubair, first nice to talk to you. Could you plz help me with how to analyze the data when ur dependent variable is ratio and ur independent variable is continous (scale) ?
how to do it ?
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 ?

▶ Reply to This

Permalink Reply by Muhammad Zubir on February 14, 2010 at 11:31pm

Dear Ismail..Before you analyse further, do check out your items and measurement. Having significant or not...it's not the answer a long as you are doing it in a proper and right manner. The finding is not on our favour, let the data and the method show what the answer of your research.

▶ Reply to This

Permalink Reply by Jean-Paul Gagnon on February 13, 2010 at 2:53am

Dear Ismael,

67 out of 920 is a difficult number to claim statistical representation for the population. As Abiodun suggested, what is key at this stage is checking how representative your 67 respondents are concerning your population. Does the sample of 67 come from roughly all areas of the peninsula? Do they represent small, medium, large, international, etc, sized firms? In essence, how well do the 67 cover your variables? If your sample of the population is good enough you may be able to stipulate some thin inferences. Unfortunately, unless the entire population of 920 responds there will be some deviance. However, as we all know, getting an entire population is nearly impossible, so the best you can do is try to get a greater number of respondents so that your error margins will be lower. Again, as Abiodun said (and which is the biggest pain in the butt with research) is follow ups on your target population. You've got to be persistent in your contact methods to get more respondents. Is it possible to offer a competition to the people who reply, like winning a gift voucher or something similar? Furthermore, who inside the individual organization needs to answer your questions? The CEO? Marketing or PR director/manager? Regular employee?

I hope this will help you a little bit. Remember to be tenacious, try to get those numbers so your study will have a greater impact.

Hoping you are well,

Jean-Paul Gagnon

▶ Reply to This

Permalink Reply by Abiodun .A. Joachim on February 13, 2010 at 11:28am

Dear Ismael,


Again, if your responses have been mainly from a geograhical area you could scaled down your geographical scope (of study) to that geographical environment and base your analysis and recommendations on that environment. Future studies on that line will learn from your difficulties and extend the study to uncovered area. Jean-Paul comments on ''who within the individual organization needs to answer your questions? is quite instructive in designing your follow up strategies. Of course, it is probable that some of your instrument may have been administered on wrong respondents in some of the organization you are studying or secretaries who promised to pass on the instrument to the right persons who never did.As you persist in your follow up see if some of those who responded now can help influence one or two of their colleagues in the industry to your side

Best of luck

▶ Reply to This

Permalink Reply by ISMAEL YOUNIS ABUJARAD on February 13, 2010 at 1:30pm

Dear Mr. Abiodun A. Joachim,
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 ?

▶ Reply to This

Permalink Reply by ISMAEL YOUNIS ABUJARAD on February 13, 2010 at 1:29pm

Dear Jean-Paul
Gagnon,
I am so grateful to your input. Well, my unit of analysis is the organization. In my questionnaire, the main parts like (organizational innovativeness) is to be responded by the CEO. However, there are other parts such as financial performance (accountign department has to answer) and other section concerning employees that the HR department has to answer. Well, it is so tough to get more than 69. This is the last number of questionnaires i managed to get. I managed to get from different states in Peninsular Malaysia. I am sorry to say that I have tried all means, but the respondents seem not interested to respond because to them it might be a waste of time. it is not a business deal. I am sorry to say that the housing developers in peninsular malaysia seem not to have social responsibility or even responsibility towards the housing industry . I am so disappointed indeed.
Do you think since I can't get enough response through which i can draw conclusions about the entire populations, i should conduct interviews and thus my study will be following a mixed method rather than quantitative ?
Do you advise me to conduct interviews ?
I even tested my model. I did not get it significant ?????????i am sad ?????????
Here 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 ?
plz advise me ?

▶ Reply to This

Permalink Reply by Jean-Paul Gagnon on February 15, 2010 at 4:18am

Dear Ismael,

First of all, do not despair. Every researcher has difficulty getting the sample they want of a distinctive population. What matters is that you evidently tried hard and got 69 respondents to draw conclusions from. There is also nothing wrong with doing a mixed methods analysis and interviews or focus groups could variably be of great use depending on how you tailor your study. Furthermore, have you thought about doing a meta-analysis? Are there other studies that are similar to yours in other countries perhaps or within different industries? If you gather studies like your and analyze them such could provide an interesting point of comparison with your quantitative results. Also, is your study with a university or for government? It's a marginal question but may help your study.

In regards to SPSS, don't mix up your DV and IV's unless you are conducting separate statistical runs with a different DV each time. It makes sense that your DV is financial profitability measured against the IV's in your study. How many questions were on the survey? Were they all 7 point Likert scales? After getting back your surveys, did you collapse your results into bigger categories or did you systematically run them against the DV in SPSS? You are trying to assess what areas or practices within the housing industry are linked more to profitability right? It's an interesting study. If you can, please answer those questions and ask some more so I can continue to get a better understanding of your work.

▶ Reply to This

• ‹ Previous
• 1
• 2
• Next ›
• Page