Sample Size Determination

[Sandeep]

Dear Group Members,

I am a Doctoral Student of Haryana School of Business, GJUS&T, Hisar. I  am studying key factors behind selection of ready-to eat foods. In my questionnaire around 70 statements are on likert scale, 20 statement are of Yes/ No type, 3 statement on multiple choice, 2 statements are open ended.  I have done my pilot survey over 210 respondents and now want to know sample size for my study. I have tried to determine the same by using SPSS in the following way:

1. Only calculated std. deviation of all the 70 statements on likert scale through SPSS
   
2. Then took mean of all std. deviation of 70 statements  = .923
3. Calculated sample size by using the formula : (Sq. of Std Dev) X (Sq. of z value) / (Sq. of level of precision)
4. I calculated sample size by using z value at 95% CL and level of precision 5 %, which come out ( As given in Marketing Research by Naresh Malhotra) = .121 which is wrong

Kindly rectify me by addressing the following queries:

1. Whether I need to calculate Std Deviation of all kinds of statements stated above?
2. To calculate the std deviation of all statements , mean method  is appropriate?
3. How to determine size of pilot survey?
4. how we can calculate the sample size by using proportion method and Can I use it in my study.

I am also attaching here with soft copy of the formula given in Marketing Research by Naresh Malhotra.

Thanking You,

Sandeep

(M) - 9996117410

 

[ishani chopra]

Dear Sandeep ji,

Since you are conducting a study on factors affecting the selection of ready to eat food and the proportion of people selecting it in the population i.e. P is unknown, you may use Cochran(1963:75) formula to calculate the sample size assuming the level of significance you desire in your study. the formula is:

n= (Z^2*p*q)/ e^2. Assuming p=0.5 as it can be assumed to be maximum variability. The resulting sample size can be calculated. I hope you would be benefited by going through the attached file on determining the sample size and especially the tables on page 3 and 4.

Regards 

--
Protocols of this Group:
 
1. Plz search previous post in group before posing the question.
2. Don't write query in someone's post. Always use the option of New topic for the new question. You can do this by writing to dataanalys...@googlegroups.com
3. Its better to give a proper subject to your post/query. It'll help others while searching.
4. Never write Open ended queries. This group intend to help research scholars NOT FOR WORK THEM.
5. Never write words like URGENT in ur posts. People will help them when they are free.
6. Never upload any info about National Seminars/Conferences. Send such info on personal emails. And feel free to share any RESEARCH related info.
7. No Happy New Year, Happy Diwali, Happy Holi, Happy B'day, Happy Anniversary etc. allowed on this group.
8. Few months back there was a facility for asking & sharing the Research Papers. Now there is no provision of asking for the research paper here.
 
Let’s make a better research environment.
---
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--

Dr. Ishani Patharia Chopra
Asst. Professor
Department of Commerce
BPSMV, Khanpur Kalan, Sonipat.

*

[Sandeep]

Dear Mam,

Thanks for the reply.

Determining the sample size by using proportion method is absolutely fine and by using the formula suggested by you, minimum sample size of 385 for my study is sufficient.

Nevertheless, if we want to calculate the sample size by using mean method or standard deviation method then I am facing lot of problems. You are right that that standard deviation is unknown to us. But to know it we either estimate it on the basis of researcher's judgment or similar studies conducted in the past or by conducting the pilot survey (As per Essentials of Marketing Research By William Zikmund, Marketing Research by Naresh Malhotra etc) . In my case I have conducted the pilot survey over 200 people, now we can calculate the standard deviation so as the sample size. Now please let me know how to calculate the standard deviation of the pilot survey by using SPSS.

Moreover to calculate the Std deviation, whether I will consider only those statement which are on likert scale or I consider all types of statements presented in the questionnaire.

Thanking You,

Sandeep

To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.

To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

[ishani chopra]

As far as my know how, Standard deviation of only metric data has some meaning. So, for calculating s.d. you may go ahead with scaling and ratio type data only.
Regards

Dr. Ishani Patharia Chopra
Assistant Professor, Department of Commerce, BPS Women University, Khanpur Kalan, Sonipat.

To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.

To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--

Dr. Ishani Patharia Chopra
Asst. Professor
Department of Commerce
BPSMV, Khanpur Kalan, Sonipat.

*

--
Protocols of this Group:
 
1. Plz search previous post in group before posing the question.
2. Don't write query in someone's post. Always use the option of New topic for the new question. You can do this by writing to dataanalys...@googlegroups.com
3. Its better to give a proper subject to your post/query. It'll help others while searching.
4. Never write Open ended queries. This group intend to help research scholars NOT FOR WORK THEM.
5. Never write words like URGENT in ur posts. People will help them when they are free.
6. Never upload any info about National Seminars/Conferences. Send such info on personal emails. And feel free to share any RESEARCH related info.
7. No Happy New Year, Happy Diwali, Happy Holi, Happy B'day, Happy Anniversary etc. allowed on this group.
8. Few months back there was a facility for asking & sharing the Research Papers. Now there is no provision of asking for the research paper here.
 
Let’s make a better research environment.
---
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.

To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.

[Sandeep]

Dear Mam,

kindly throw some light on how to calculate Standard Deviation through SPSS. Earlier, I calculated the std Deviation of all individual statements, on Likert Scale, by using SPSS and then took the mean of all the std deviations in excel sheet.

Is it the right way to calculate the Std Deviation for the formula given in Marketing Research by Naresh Malhotra?

Thanking You,

Sandeep

Assistant Professor

Kedarnath Aggarwal Institute of Management

(M) - 9996117410
Email: sande...@kaimindia.in

 

You received this message because you are subscribed to a topic in the Google Groups "DataAnalysis" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/dataanalysistraining/t8ouPKjb3GM/unsubscribe.
To unsubscribe from this group and all its topics, send an email to dataanalysistrai...@googlegroups.com.

[Chirag Malik]

Dear Sandeep,

As practiced by so many researchers in India and discussed in data analysis workshops, (and I am sure you would also be knowing) the rule of thumb for sample size determination is (no. of statements)X10. I do not find any reason for not accepting this rule of thumb in your case as well.

Now lets talk about the systematic approach to find sample size using standard deviation(SD). As value of Z and ME or level of precision is decided by you only, it is the SD only, which will decide the sample size. To calculate the SD of the population from which you have planned drawing sample, you must first understand for which particular characteristic you are looking for the variance in the population. Now, since your seventy statements would definitely measuring one particular factor/event/dimension/characteristic  of the population (I am sure that the internal consistency of the pilot data is well taken care of), so you have to calculate the SD of that particular characteristic.

Now, how to do it on SPSS? Take your pilot survey data, take the summated value of all the statements for each respondent. Divide each value by 70. You will get average score of each of the respondents for all the statements. This is the value of the factor/dimension for each respondent which you are measuring through these 70 statements. Now use 'descriptive' tab and take the SD of these values. Now apply the formula, I am sure you would not be disappointed.

Happy learning,

Chirag Malik

Asst. Professor,

GIBS, Rohini, New Delhi.

On Sat, Oct 10, 2015 at 3:02 PM, ishani chopra <isha...@gmail.com> wrote:

[Rupali Madan]

Sir
As u have explained the thumb rule is no.of statements multiplied by 10.
Suppose a questionnaire is containing 70 questions then sample size should be 700.
I m not able to understand thus logic.

[Chirag Malik]

Dear Ms Rupali,

As I have also mentioned that some researchers use it and it has also been discussed in various data analysis workshops that it is taken as 'rule of thumb'. 

Please understand the meaning of 'rule of thumb'. It means that if you do not want to go by some systematic method or long route of doing the number of calculations and evaluations, then you can always apply this rule. 

As far as logic is concerned, it has been found that in the majority of cases it fits the requirement of minimum sample size.

However, one may always follow the systematic way of calculating the sample size. 

Happy learning, 

Chirag Malik

Asst. Professor

GIBS, Rohini, New Delhi.

[ishani chopra]

Yes. It is all about adequacy of data which is one of the necessary conditions for Factor Analysis. 
But there are various approaches to determination of sample dize. The objective and subjective factors related to research must be taken care of. Larger the sample better it is.
With regard to Cochran formula, you may increase the sample size depending upon the desired precision level and population. . 

Dr. Ishani Patharia Chopra
Assistant Professor, Department of Commerce, BPS Women University, Khanpur Kalan, Sonipat.