Depending upon which statistical variable is to be found out for the main study, the sample size for itself vary. Most studies fail to estimate the pre-study sample size for the main study so sample size for pilot study is be estimated by the author on feasability and results of the pilot should be calculated with caution. Therefore, half  or quarter of the actual sample size calculated with the general formula {p*q(z2)/(0.05)2 }  for the main study should be about right.   

about more for pilot study , you can go for this link

http://www.biomedcentral.com/content/pdf/1471-2288-10-67.pdf&embedded=true

Based on study objectives, the sample size has to be determined. In case of categorical analysis (say effect of socioeconomic characteristics on the responses), each category should have a satisfactory sample size. The required sample size could also be varied based on the sampling technique, population proportion, degree of accuracy, confidence interval.

For more detail refer the book "Research Methodology by C. R Kothari 2014"

You can calculate your sample size depending on the statistical test you want to use/ parameter you want to analyze, the expected effect size, the minimal power and the alpha error level. If you want to use tests like t-test, ANOVA or multiple regression, there is this software GPower which deals with this (it is open source, just search it on google, there should be a link to university of duesseldorf). When you want to do more complex statistical tests you should search for a paper on sample size calculations for these tests.

Anil -

The number of questions does not impact the sample size needs, which vary by question, except that too much respondent burden may increase bias, and variance.  Sample size needs are based on population standard deviation, complicated by type of data, type of design, use of auxiliary data, if any, and assume you have negligible bias, which may be a bad assumption.  Standard error for a statistic such as a mean is basically then determined by population standard deviation and sample size.  You have to decide what standard errors you need.  (This is for quantitative data.  For qualitative data - not my area - I suppose the idea of low bias ("representativeness"), and sufficient sample size to account for variability, is still the case, in an analogous way.) 

The standard error goals determine what you need, and there may need to be a compromise, depending upon which questions are more important, and what is feasible.  A pilot study can be used to estimate population (or subpopulation/stratum) standard deviation.  Knowing when to stop may be tricky.  You just have to do something reasonable.  The better the estimate of standard deviation, the better the estimate of sample size requirement.  You might also consider similar populations.  I know that Cochran(1977) below, and perhaps others, have such suggestions. 

Standard errors are important for confidence intervals, and in hypothesis tests. 

If you use this for hypothesis testing, you are looking at type II and type I error probabilities together.  The type II error probability relates to "power." 

Careful of interpreting hypothesis tests.  P-values do not stand alone. 

Here are some books that may be helpful:

Cochran, W.G(1977), Sampling Techniques, 3rd ed., John Wiley & Sons. 

Blair, E. and Blair, J(2015), Applied Survey Sampling, Sage Publications.

Lohr, S.L(2010), Sampling: Design and Analysis, 2nd ed., Brooks/Cole.

Särndal, CE, Swensson, B. and Wretman, J. (1992), Model Assisted Survey Sampling, Springer-Verlang.     

Brewer, KRW (2002), Combined survey sampling inference: Weighing Basu's elephants, Arnold: London and Oxford University Press.  

An Introduction to Model-Based Survey Sampling with Applications, 2012,

Ray Chambers and Robert Clark,  Oxford Statistical Science Series

Finite Population Sampling and Inference: A Prediction Approach,  2000,

Richard Valliant, Alan H. Dorfman, Richard M. Royall

I think that one of the first three would be more likely to be best to obtain first.  That may be all you need. 

I hope I have not misunderstood the emphasis of your question. 

Cheers - Jim 

PS -  Note that many online sample size "calculators," tacitly assume simple random sampling for proportions, assuming the worst case rather than estimate standard deviation, and generally assume no need for a finite population correction  (fpc) factor, which may sometimes calculate a sample size need that is larger than the population size.  Beware. 

PSS - Oversimplified summary, for each question:

A pilot study is one way to estimate standard deviation.  If you can randomly ignore some of your pilot data and still get about the same standard deviation estimate, you are fine. 

Whatever sample size, along with that standard deviation, will give you an acceptable standard error, is going to be what you want.  

Dear All

In order to answer this question/problem, several remarks have to be studied.

1. General remarks:

Research studies are usually carried out on sample of subjects rather than whole populations. The most challenging aspect of fieldwork is drawing a random sample from the target population to which the results of the study would be generalized. The key to a good sample is that it has to be typical of the population from which it is drawn. When the information from a sample is not typical of that in the population in a systematic way, we say that error has occurred. In actual practice, the task is so difficult that several types of errors, i.e. sampling error, non-sampling error, Response error, Processing error,…

In addition, the most important error is the Sampling error, which is statistically defined as the error caused by observing a sample instead of the whole population. The underlying principle that must be followed if we are to have any hope of making inferences from a sample to a population is that the sample be representative of that population. A key way of achieving this is through the use of “randomization”. There several types of random samples, Some of which are: Simple Random Sampling, Stratified Random Sampling, Double-stage Random Sampling... Moreover, the most important sample is the simple random sample which is a sample selected in such a way that every possible sample of the same size is equally likely to be chosen. In order to reduce the sampling error, the simple random sample technique and a large sample size have to be developed.

2. Specific remarks:

The following factors are highly affected the sample size and need to be identified:

• Population Size,

• Margin of Error,

• Confidence Level (level of significance) and

• Standard of Deviation.

Then, the sample size can be estimated by,

Necessary Sample Size = (z-score or t-value)2 * StdDev*(1-StdDev) / (margin of error)2 .

In addition to my previous answer, I would like to mention that there several types of random samples, Some of which are: Simple Random Sampling, Stratified Random Sampling, Double-stage Random Sampling... Moreover, the most important sample is the simple random sample which is a sample selected in such a way that every possible sample of the same size is equally likely to be chosen. In order to reduce the sampling error, the simple random sample technique and a large sample size have to be developed.

Very useful replies. Thank you.

Useful contributions! Thank you.