See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/322887480

Determining Sample Size; How to Calculate Survey Sample Size

Article · February 2017

CITATIONS READS

15 60,551

1 author:

Hamed Taherdoost
Hamta Group
93 PUBLICATIONS   719 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Hamed Taherdoost on 02 February 2018.

The user has requested enhancement of the downloaded file.

 International Journal of Economics and Management Systems
Hamed Taherdoost http://www.iaras.org/iaras/journals/ijems

Determining Sample Size;

How to Calculate Survey Sample Size
HAMED TAHERDOOST a, b
a
Research and Development Department, Hamta Business Solution Sdn Bhd, Malaysia
b
Research and Development Department, Ahoora Ltd | Management Consultation Group, Malaysia
[email protected] http://www.ahooraltd.com http://www.hamta.org

Abstract: - The sample size is a significant feature of any empirical study in which the goal is to make inferences
about a population from a sample. In order to generalize from a random sample and avoid sampling errors or
biases, a random sample needs to be of adequate size. This study presents a summary of how to calculate the
survey sample size in social research and information system research.

Key-Words: - Sample Size, Survey Design, Questionnaire Development, Calculate Sample Size, and Research
Methodology

1 Survey Sample Size There are two key factors to this formula (Bartlett et
al., 2001). First, there are considerations relating to
the estimation of the levels of precision and risk that
In order to generalize from a random sample and
the researcher is willing to accept:
avoid sampling errors or biases, a random sample
E is the margin of error (the level of precision) or the
needs to be of adequate size. What is adequate
risk the researcher is willing to accept (for example,
depends on several issues which often confuse people
the plus or minus figure reported in newspaper poll
doing surveys for the first time. This is because what
results). In the social research a 5% margin of error
is important here is not the proportion of the research
is acceptable. So, for example, if in a survey on job
population that gets sampled, but the absolute size of
satisfaction 40% of respondents indicated they were
the sample selected relative to the complexity of the
dissatisfied would lie between 35% and 45%. The
population, the aims of the researcher and the kinds
smaller the value of E the greater the sample size
of statistical manipulation that will be used in data
required as technically speaking sample error is
analysis (Taherdoost, 2016). While the larger the
inversely proportional to the square root of n,
sample the lesser the likelihood that findings will be
however, a large sample cannot guarantee precision
biased does hold, diminishing returns can quickly set
(Bryman and Bell, 2003).
in when samples get over a specific size which need
Z concern the level of confidence that the results
to be balanced against the researcher’s resources
revealed by the survey findings are accurate. What
(Gill et al., 2010).
this means is the degree to which we can be sure the
To put it bluntly, larger sample sizes reduce sampling
characteristics of the population have been accurately
error but at a decreasing rate. Several statistical
estimated by the sample survey. Z is the statistical
formulas are available for determining sample size.
value corresponding to level of confidence required.
There are numerous approaches, incorporating a
The key idea behind this is that if a population were
number of different formulas, for calculating the
to be sampled repeatedly the average value of a
sample size for categorical data.
variable or question obtained would be equal to the
true population value. In management research the
𝑝 (100−𝑝)𝑧 2 typical levels of confidence used are 95 percent
𝑛 =
𝐸2 (0.05: a Z value equal to 1.96) or 99 percent (0.01:
Z=2.57). A 95 percent level of confidence implies
n is the required sample size that 95 out of 100 samples will have the true
P is the percentage occurrence of a state or condition population value within the margin of error (E)
E is the percentage maximum error required specified.
Z is the value corresponding to level of confidence The second key component of a sample size formula
required concerns the estimation of the variance or
heterogeneity of the population (P). Management

ISSN: 2367-8925 237 Volume 2, 2017

 International Journal of Economics and Management Systems
Hamed Taherdoost http://www.iaras.org/iaras/journals/ijems

researchers are commonly concerned with ` Variance of the population P=50%

determining sample size for issues involving the
estimation of population percentages or proportions Confidence level=95%
(Zikmund, 2002). In the formula, the variance of a Margin of error
proportion or the percentage occurrence of how a 25000 378 1023 6939
particular question, for example, will be answered is 50000 381 1045 8057
P(100-P). Where, P= the percentage of a sample 100000 383 1056 8762
having a characteristic , for example, the 40 % of the 250000 384 1063 9249
respondents who were dissatisfied with pay, and 500000 384 1065 9423
(100-P) is the percentage (60%) who lack the 1000000 384 1066 9513
characteristic or belief. The key issue is how to
estimate the value of P before conducting the survey?
Bartlett et al. (2001) suggest that researchers should
use 50% as an estimate of P, as this will result in the Table 2: Sample size based on Desired Accuracy
maximization of variance and produce the maximum with Confidence Level of 99%
sample size (Bartlett et al., 2001). Source: (Gill et al., 2010)
The formula for determining sample size, of the
population has virtually no effect on how well the ` Variance of the population
sample is likely to describe the population and as P=50%
Fowler (2002) argues, it is most unusual for it (the Confidence level=99%
population fraction) to be an important consideration Margin of error
when deciding on sample size (Fowler, 2002). Population Size 5 3 1
Table 1 and 2 present sample size that would be 50 46 49 50
necessary for given combinations of precision, 75 67 72 75
confidence levels, and a population percentage or 100 87 95 99
variability of 50% (the figure which many 150 122 139 149
researchers suggest to maximize variance). 200 154 180 198
250 181 220 246
Table 1: Sample size based on Desired Accuracy 300 206 258 295
with Confidence Level of 95% 400 249 328 391
Source: (Gill et al., 2010) 500 285 393 485
600 314 452 579
` 700 340 507 672
Variance of the population P=50%
800 362 557 763
Confidence level=95% 1000 398 647 943
Margin of error 1500 459 825 1375
Population Size 5 3 1 2000 497 957 1784
50 44 48 50 3000 541 1138 2539
75 63 70 74 5000 583 1342 3838
100 79 91 99 10000 620 1550 6228
150 108 132 148 25000 643 1709 9944
200 132 168 196 50000 652 1770 12413
250 151 203 244 100000 656 1802 14172
300 168 234 291 250000 659 1821 15489
400 196 291 384 500000 660 1828 15984
500 217 340 475 1000000 660 1831 16244
600 234 384 565
700 248 423 652
800 260 457 738
1000 278 516 906
2 Conclusion
1500 306 624 1297
As mentioned, for sample size calculation, Table 1
and 2 or the provided formula can be used. The
2000 322 696 1655
sample sizes reflect the number of obtained
3000 341 787 2286
responses, and not necessarily the number of
5000 357 879 3288
questionnaires distributed (this number is often
10000 370 964 4899

ISSN: 2367-8925 238 Volume 2, 2017

 International Journal of Economics and Management Systems
Hamed Taherdoost http://www.iaras.org/iaras/journals/ijems

increased to compensate for non-response).

However, in most social and management surveys,
the response rates for postal and e-mailed surveys are
very rarely 100% (Taherdoost, 2016). Probably the
most common and time effective way to ensure
minimum samples are met is to increase the sample
size by up to 50% in the first distribution of the
survey (Bartlett et al., 2001).

References:

[1] BARTLETT, J. E., KOTRLIK, J. W. & HIGGINS,

C. C. 2001. Organizational Research: Determining
Appropriate Sample Size in Survey Research.
Learning and Performance Journal, 19, 43-50.
[2] BRYMAN, A. & BELL, E. 2003. Business Research
Methods, Oxford, Oxford University Press.
[3] FOWLER, F. J. 2002. Survey research methods,
Newbury Park, CA, SAGE.
[4] GILL, J., JOHNSON, P. & CLARK, M. 2010.
Research Methods for Managers, SAGE
Publications.
[5] TAHERDOOST, H. 2016. Sampling Methods in
Research Methodology; How to Choose a Sampling
Technique for Research. International Journal of
Advance Research in Management, 5(2), 18-27.
[6] TAHERDOOST, H. 2016. How to Design and Create
an Effective Survey/Questionnaire; A Step by Step
Guide. International Journal of Advance Research in
Management, 5(4), 37-41.
[7] ZIKMUND. 2002. Business Research Methods,
Dryden, Thomson Learning.

ISSN: 2367-8925 239 Volume 2, 2017

View publication stats