Practical factors

As with much else in organizational life it is often the practical considerations that, in the end, dominate many decisions. Cost, time and the availability of suitable personnel cannot be ignored. Cost is often the dominant factor in determining how many interviews are undertaken.

For all the reasons given above, cost is not a relevant determination of sample size, but this ignores practical reality. Some trade-off must be achieved between increased reliability and accuracy in the data arising from a larger size sample and increased cost, so that an optimum level is achieved. If cost is allowed to be the determinant of sample size then the implications for reliability and accuracy must be recognized and not ignored as is so often the case. It cannot be expected that the same quality of results will be achieved from a small sample as from a large sample. However, it is unnecessary to generate a high level of precision in research results when other areas surrounding the problem under consideration are subject to an even higher degree of error. Time is another important element in that the shorter the time in which results are required, then the less time available for fieldwork and the smaller the size of sample that can be achieved in the time limits set. Finally, suitably trained interviewers may simply not be available to meet the needs of a very large and widely spread sampling requirement.

A number of factors have been shown to be relevant in determining sample size. This demonstrates that the appropriate sample size must be worked out with respect to the needs of each particular survey. However, most beginners in a subject area like some rough clues as to the order of size that may be 'usual'. As 'rule of thumb' guidelines only, the following points may be helpful. In general, for acceptable statistical validity of results generated from quantitative surveys, any subgroup containing fewer than 100 respondents should be treated with extreme caution in statistical analysis. Numbers below 50 respondents should not be subjected to statistical analysis at all. The normal range of sample sizes used in national samples for many consumer goods is 1500-2000 respondents. Minimum sample sizes for quantitative consumer surveys are in the order of 300-500 respondents. The upper limits of size of sample have been pushed very high by the Government's National Housing Survey including 100,000 households, but this is most unusual.

The points made above with respect to sample size have been derived from sampling theory, which forms the basis for random sampling. In the case of quota sampling, or of variations to the random procedure such as multi-stage and random location, it is usual to increase the size of the sample to compensate for any inaccuracies introduced by the sampling procedure. Alternatively, a 'design factor' may be used to reduce the confidence level and accuracy in the results, and compensate for the limitations of the non-random sampling techniques in this way.

As a final note on sampling, it should be mentioned that the only time when the size of the underlying population needs be taken into account in considering the size of the sample to be drawn from it, is when the size of the sample required is likely to account for 10 per cent or more of the population.


Online Survey Champion

Online Survey Champion

There are people all over the world trying to find ways to make money online. From stay at home moms looking to make a few extra dollars to college students and entrepreneurs, the allure of making your own hours and working from home or from the local coffee shop is very appealing.

Get My Free Ebook

Post a comment