Home › Forums › Default Forum › Quantitative study – sample size issue
- This topic has 13 replies, 6 voices, and was last updated 10 years, 3 months ago by
Shyh-Mee Tan.
-
AuthorPosts
-
19th October 2010 at 1:57 am #3983
roshayu mohamad
MemberHi,
Does anyone have a good reference on this?
My problem:
I am conducting a quan study on 5 public univs in Malaysia (out of 20 public unis in total)
Sampling frame: public higher education administrative staff and managers.
population size: 50,000.
Thank you.
19th October 2010 at 2:01 am #3996Shyh-Mee Tan
Participant19th October 2010 at 2:32 am #3995Kerry MacQuarrie
MemberI responded to your post in the quantitative methods group but seeing this here, will ask even more questions. Namely, could you tell us more about the qualitative component of your research (research question, data collection methods, how much of the study instruments will be quali, etc) and how you are combining qualitative & quantitative in your study?
One rule of thumb in quali research is to keep sampling until you reach saturation, which is the point where you cease to gather NEW information. This leads sampling to be an iterative, open-ended endeavor, much to the ire of the funding proposal developer, project budget manager or field logistics coordinator.
On the other hand, much quantitative data is gathered to maximize statistical significance, minimize standard errors, and ensuring “large enough” cell sizes in whatever models/with whichever variables you’re running, keeping in mind potential loss to follow-up, non-response, skip patterns or response categories you may eliminate, and other restrictions to your analytic sample.
Both are different approaches at achieving a similar goal: to be able to say that the information you get from your data is somehow “true” of your study population and not some anomaly based on the uniqueness of the individuals selected. Either way, there is no quick an easy answer to the question “How many respondents is enough?”.
19th October 2010 at 2:49 am #3994Dr Razali Hassan
MemberRoshayu…
Small matter.. what is your ‘real’ problem? Specifies your sampling type. Purposive or random or….? Should be no problem.
Cheers.Rgds
19th October 2010 at 5:00 pm #3993Shyh-Mee Tan
ParticipantDear all,
Mine not actually a problem. I did stratified samplings. The number of samples that I should have are about 335 samples. The number that I should sent out is 650 samples. During my data collections, I managed to get 710 samples.
Therefore, will there be any issues on return rates? Can I just report the number of samples that I get?
19th October 2010 at 11:40 pm #3992roshayu mohamad
MemberThanks Shyh Mee. I found that useful.
19th October 2010 at 11:44 pm #3991roshayu mohamad
MemberDr Razali,
I am using purposive sampling.. but with 20 univ in Malaysia, I only manage to get 5 to participate in the study. Would that be sufficient?
Data from MOHE shows that there are 50,000 non-academic staff in public Uni – so far I am using Krejecie & Morgan table – with that amount of population, sample size should be around 380.
19th October 2010 at 11:47 pm #3990roshayu mohamad
MemberDave,
For now I am using the Krejecie and Morgan table to guide my number of participants that I should recruit.
However, given the big population, just wonder is 5 out of 20 public university is reasonable for the study.
19th October 2010 at 11:55 pm #3989roshayu mohamad
MemberKerry,
I am studying cultural factors that affect knowlege management practices in higher education.
The qual part is used to gather this factors from the employees perspective (using interview) , and use these result to develop the quantitative instrument that is the quustionnaire.
I’ve completed the interview until I reach the saturation state. And now recruiting participants for the quan stage.
I am using Krejecie and Morgan table for the quan sample size – but just wonder giving the big population of univ (20) is 5 univ enough to be included in the study.
Thanks.
20th October 2010 at 2:37 am #3988Shyh-Mee Tan
ParticipantDear Dave,
Thanks. I appreciate your comments.
Shyhmee
20th October 2010 at 11:56 am #3987Dr Razali Hassan
MemberRoshayu,
Do you think finding from 5 uni can be generealise to others uni? What is your focus point and who is your purposive respondent?
25th November 2010 at 2:26 am #3986Krishamurthy Prabhakar
MemberThanks for your good comments. I am sharing your thoughts with my students. Dr.K.Prabhakar
25th November 2010 at 2:29 am #3985Krishamurthy Prabhakar
MemberDear Dave, Greetings from India. Many times i found it difficult to teach the concept of effect size to students of business studies. Is it possible to help me? I need a simple explanation that will help students to appreciate the concept of effect size.
29th December 2010 at 12:45 pm #3984Frederick Forson
Memberread Cresswell (2002) & Bartlett, J.E; Kotrlik, J.W; Higgins, C. C.
(2001).Organizational research: determining appropriate
sample size in survey research (Electronic version). Information technology learning
and performance journal. Spring.
vol. 19,No.1 spring. Retrieved 13/04/2010 at 12:05
noon, from http://http://www.bartlettkotrlikhiggins.pdf.adobe
. -
AuthorPosts
- The forum ‘Default Forum’ is closed to new topics and replies.