Code and Title of the Paper: F15RM Research Methods in Nutrition

Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

F15RM07 Sampling techniques – sample size and its determinations

Quadrant - I

Introduction
Selection of sample which is representative of the entire population plays a
pivotal role in applied statistics research. Sampling technique and determination of
sample size have a crucial role in research problems. Sampling technique used in a
research problem must be specific and appropriate for a particular target problem as
one technique may not help to derive solution for all problems. One important
question in undertaking a research is that what should be the size of the sample? Or
how big or small a sample must be taken for a good survey? Hence, sample size
similar to a sampling technique must be appropriate else may lead to misleading
conclusions. Generally, the sample size must not be either too large or too small but
must be of optimum size. Also certain hidden bias may also lead to erroneous
outcomes. Several factors are to be considered for determining the sample size like
size of the study population, confidence level, expected proportion and standard
deviation of the outcome variable, margin of accuracy etc. There are distinct
methods for calculating sample size for different study designs and different outcome
measures.

Objectives
After completing this module, the learner would be able to
 Understand what sample size is and its significance in a research
investigation
 Arrive at the size of sample for different sampling designs
 Determine the sample size based on different parameters

Sample size
Sample size can be simply defined as the number of people to be picked to carry out
a research problem and is represented by ‘n’. It is the technique of electing the
number of observations to be included in the sample. The main objective is to obtain
Code and Title of the Paper: F15RM Research Methods in Nutrition
Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

both a desirable accuracy and a desirable confidence level of desired width with
minimum cost. The fixation of sample size will help to draw valid and generalised
conclusions. Determination of sample size varies with sampling method used. For
example
 In a survey sampling when the population involved is heterogeneous stratified
sampling, sampling size would be different for each population.
 In census, the sample size and population size must be same.
 In experimental study, where different experimental subgroups are present,
the sample size must be determined for each experimental group.
Generally sample size determination may be as follows:
 Cost base – items which are readily available and easy to collect must be
chosen. Small sample size might result in wide confidence intervals or errors
in testing the statistical hypothesis.
 Variance base - Using a target variance for an estimate to be derived from
the sample eventually obtained
 Statistical power base - Using a target for the power of a statistical test to be
applied once the sample is collected. Sample sizes are judged based on the
quality of the resulting estimates, sample size may be assessed based on the
power of a hypothesis test.
Sample size determination
Size of the sample must be determined based on the following criteria
1. Nature of the universe – the nature of the universe may be homogenous or
heterogenous. If the nature of the universe is homogenous, a small sample
can be used. If the nature of the universe is heterogenous a large sample may
serve the purpose and is termed as dispersion factor technically.
2. Number of classes proposed – if there are many classes like groups and
sub groups are formed large sample solves the purpose as small sample
would not be a representative of the number of items in each sub group.
3. Nature of the study – for intensive and continuous study and for technical
surveys, small sample is the best. But a large sample is required for general
survey.
Code and Title of the Paper: F15RM Research Methods in Nutrition
Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

4. Type of sampling – sampling technique plays an important role in selecting

the size of the sample. A small random sample is apt to be much superior to a
larger but badly selected sample.
5. Standard of accuracy and acceptable confidence level – in order to keep
high the standard of accuracy and precision level a relatively larger sample is
required. To increase the accuracy double the times, the sample size must be
increased fourfold.
6. Sampling error – increase in sample size may reduce the sampling error.
Sampling error becomes zero if the size of the complete population equals the
sample size. But increase in sample size increases cost, time and efforts in
data collection.
7. Availability of finance – the amount of money available to conduct the study
is also important and depends on the size of the sample. Finance should be
kept in mind while determining the sample size as larger samples may
increase the cost estimates.
8. Other considerations – the nature of units, population size, length of the
questionnaire, availability of trained investigators, the sample conditions, time
frame of the study are certain considerations to which a researcher must pay
attention while selecting the size of the sample.
Significance of sample size
In a comparative study, comparison between two groups is obtained by the
means or proportions of a particular characteristic. To determine the significance
between or within the groups a statistical test is applied on the means or proportions
observed in the different groups. Here, sample size principally has an effect on the
statistical power, which is nothing but a statistical test to indicate the level of
significance if truly present. If the statistical test demonstrates no significant
difference then issue arises. If such negative result occurs there may be only two
possibilities. One interpretation is that the results of the statistical test is right and no
significant difference statistically i.e. a true negative result. Other interpretation is
that the statistical test may be erroneous with actually a difference for which there
was no powerful sample size to identify the different i.e. false negative result also
Code and Title of the Paper: F15RM Research Methods in Nutrition
Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

known as Type II error statistically. So adequate sample size is required to prove the
significant high statistical power. Low statistical power may rise due to inadequate
sample size.
Determination of the sample size for parameters
The sample size generally depends on five parameters
1. Basis of Minimum Expected Difference or size effects
The statistical significance increases when the sample size is large with a less
minimum expected difference. For example, if a study is designed to compare a
standard diagnostic procedure of 85% accuracy with a new procedure of unknown
but potentially higher accuracy and the investigator believes that it would be a
clinically important improvement if the new procedure were 95% accurate. Then the
investigator would choose a minimum expected difference of 10% (0.10).
2. Estimated Measurement Variability
This parameter is represented by the expected standard deviation in the
measurements decided within each comparison group. Statistical variability and
sample size needed to detect the minimum difference are directly proportional to
each other. The estimated measurement variability should be determined on the
basis of preliminary data collected from a similar study population. A review of the
literature can also provide estimates of this parameter. If preliminary data are not
available, the parameter can be estimated on the basis of subjective experience, or a
range of values may be assumed. A separate estimate of measurement variability is
not required when the measurement being compared is a proportion, as the standard
deviation is mathematically derived from the proportion.
3. Based on Statistical Power
To increase statistical power, sample size must be increased.
4. Significance Criterion (P – Value)
Maximum P value is considered to be statistically significant. As the significance
criterion is decreased, the sample size needed to detect the minimum difference
increases. The statistical significance criterion is customarily set to 5 percent.
5. One- or Two-tailed Statistical Analysis
Code and Title of the Paper: F15RM Research Methods in Nutrition
Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

In certain cases, the difference between comparison or experimental groups is

possible in only one direction. In such cases, one-tailed statistical analysis, which
require a smaller sample size for detection of the minimum difference is used than
would a two-tailed analysis. The sample size of a one-tailed study design with a
given statistical significance criterion—for example, α—is equal to the sample size of
a two-tailed design with a significance criterion of 2α, all other parameters being
equal.
Criteria for good Sample Size - to determine the appropriate sample size: the level
of precision, the level of confidence or risk, and the degree of variability in the
attributes being measured is required.
1. Level of Precision
The level of precision also known as sampling error, is the range in which the true
value of the population is estimated to be and range is expressed in terms of
percentage points (e.g., ±5 per cent). For example if a researcher finds that 60% of
farmers in the sample have adopted a recommended practice with a precision rate of
±5%, then he or she can conclude that between 55% and 65% of farmers in the
population have adopted the practice.
2. Confidence Level
The risk level of confidence level is based on ideas of Central Limit Theorem. The
key idea is that when a population is repeatedly sampled, the average value of the
attribute obtained by those samples is equal to the true population value. The values
obtained by these samples are normally distributed about the true value, with some
samples having a higher value and some obtaining a lower value than the true
population value. In a normal distribution, approximately 95% of the sample values
are within two standard deviations of the true population value. This confidence
interval is also known as risk of error in the statistical hypothesis testing. In other
words, this means that if a 95% confidence level is selected, 95 out of 100 samples
will have the true population value within the range of precision specified. There is
always a probability that the sample obtained by the researcher or investigator does
not represent the true population value. Such samples with extreme values are
Code and Title of the Paper: F15RM Research Methods in Nutrition
Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

represented. This risk is reduced for 99% confidence levels and increased for 90%
or lower levels of confidence.
3. Degree of Variability
The degree of variability in the attributes refers to the distribution of attributes in the
population. The sample size required is smaller if the variables are of homogeneous
population. The sample size required is larger to obtain a given level of precision for
a heterogeneous population.
Let we see few examples on determination of sample size
Example 1 - When project is sanctioned, the total budget is allocated for the conduct
of the field survey which is a part of the project proposal. It is easy to obtain a
sample size if the cost of sampling per sample unit is known. By dividing the total
budget allocated by the cost of the sampling per unit sample size can be determined.
In this example, only the cost aspect of the sampling is taken into account instead of
the value of information obtained from each sample.
Example 2 – few researchers decide the sample size based on previous similar
research studies. Still this is not a substitute for a formal scientific approach.
Example 3 -
Conclusion
Thus sample size plays an important role in any applied and scientific
research investigation. Sample size calculation is an essential part of a study
protocol or design. Researchers with the help of a statistician can bring about best
results of a study design. Sample size act as a centrepiece of a statistical analysis
and reflects the width of the confidence interval rather than the outcome of the test of
significance. Sample size must be adequate enough to detect even a small
difference significantly between the groups under study. If sample size is inadequate
the power of the study becomes less with decrease in precision of results. Hence
attention to sample size will result in a more authentic reliable results and
interpretation with high statistical significance.

https://en.wikipedia.org/wiki/Sample_size_determination
Code and Title of the Paper: F15RM Research Methods in Nutrition
Code and Title of the Module F15RM07 Sampling techniques – sample size and its
determinations
Name of the Content Writer: Dr.R.Balasasirekha

http://www.thejhs.org/article.asp?issn=2468-
6360;year=2014;volume=2;issue=4;spage=142;epage=147;aulast=Omair