SPEARMAN'S RANK CORRELATION Population and Sample
A Spearman rank correlation describes the Population
monotonic relationship between 2 variables. It is (1)
The totality of elements from which data
useful for nonnormally distributed continuous data,
under consideration shall be obtained.
(2) can be used for ordinal data, and (3) is relatively
robust to outliers. In statistics, population is the entire set of
items from which data is drawn for a statistical study.
It can be a group of individuals, a set of items etc. It
constitutes the data pool for a study
Sample
A representative portion of the population
taken for a study.
A population commonly contains too many
Spearman's rank correlation coefficient is a individuals to study conveniently, so an investigation
statistical measure to show the strength of a is often restricted to one or more samples drawn
relationship between two variables. Ten is the from it. A well chosen sample will contain most of the
minimum number needed in a sample for the information about a particular population parameter
spearman's rank test to be valid. but the relation between the sample and the
The Spearman rank correlation is the population must be such as to allow true inferences
nonparametric counterpart of the Pearson to be made about a population from that sample.
correlation. A sample represents the group of interest
Spearman's rank correlation examines the from the population which we will use to represent
relationship between two variables. our data. The sample is an unbiased subset of the
population which best represents the whole data
The Spearman rank correlation is the
nonparametric counterpart of the Pearson Population data is used when the data pool
correlation. is small and cooperative to giving all the required
information. For larger populations, we can use
INTERPRETATION sampling to represent parts of the population it is
A very high value of r, indicative if a very high hard to collect data from.
positive correlation between the scores in English Samples are used when the population is
and Science. It can be inferred from these results large in size, scattered or if it's generally hard to
that achievement in English is positively related to collect data on individual instances within it. We can
THE achievement in Science. Those who learned then use a small sample of the population to make
English well is likely to perform well in Science. overall hypotheses
Determining Sample Size
One factor that must be considered in
determining the sample size is the error which one
allow in making decision. Such error is referred to as
level of significance denoted by a
SAMPLING PROCEDURES
Let N be the population size and n be the sample Sampling is a method that allows
size: Slovin's formula researchers to infer information about a population
based on results from a subset of the population,
without having to investigate every individual.
Reducing the number of individuals in a
study reduces the cost and workload, and may make
it easier to obtain high-quality information, but this
has to be balanced against having a large enough
sample size with enough power to detect a true
association.
• The maximum value that could be assigned to e is
Probability Sampling
0.5 or 5%. If greater degree of accuracy is desired,
0.01 or 1%. Simple Random Sampling
EXAMPLE In this case each individual is chosen entirely
by chance and each member of the equal chance,
It is desired to conduct a survey on 5, 000
population has an probability, of being selected. One
students of National University Fairview. What is the
way of obtaining a random sample is to give each
minimum sample size maybe drawn if an error is at
individual in a population a number, and then use a
most 5% is allowed?
table of random numbers to decide which individuals
TRY to include.
There are 2,000 housewives in Fourthville EXAMPLE
Subdivision. It is desired to make a survey if their
If you have a sampling frame of 1000
preferences for cooking fuel. How many housewives
individuals, labelled 0 to 999, use groups of three
may at least be taken as respondents if an error of
digits from the random number table to pick your
at most 5% in the decision may be allowed.
sample. So, if the first three numbers from the
random number table were 094, select the individual
labelled "94", and so on.
Systematic Sampling
Systematic sampling is often more
convenient than simple random sampling, and it is
easy to administer. However, it may also lead to bias,
for example if there are underlying patterns in the
order of the individuals in the sampling frame, such
that the sampling technique coincides with the
periodicity of the underlying pattern.
This can be done by listing all the elements
in the population and selecting every kth element in
your population list. This is equally precise as the
simple random sampling. It is often used on long
population lists. To determine the interval to be used
in identifying the samples to who will participate in
the study, use the formula K = (population/sample
size).
SAMPLING EXAMPLE
Individuals are selected at regular intervals from the representative of other characteristics that weren't
sampling frame. The intervals are chosen to ensure considered (a consequence of the non- random
an adequate sample size. If you need a sample nature of sampling).
size n from a population of size x, you should select
Purposive Sampling
every x/nth individual for the sample. For example,
if you wanted a sample size of 100 from a Selective, or subjective sampling, this
population of 1000, select every 1000/100 = 10th technique relies on the judgement of the researcher
member of the sampling frame. when choosing who to ask to participate.
Researchers may implicitly thus choose a
Stratified Sampling
"representative" sample to suit their needs, or
In this method, the population is first divided specifically approach individuals with certain
into subgroups (or strata) who all share a similar characteristics.
characteristic. It is used when we might reasonably
This approach is often used by the media
expect the measurement of interest to vary between
when canvassing the public for opinions and in
the different subgroups, and we want to
qualitative research.
representation from all the ensure subgroups.
Snowball Sampling
EXAMPLE
This method is commonly used in social
In a study of stroke outcomes, we may
sciences when investigating hard-to- reach groups.
stratify the population by sex, to ensure equal
Existing subjects are asked to nominate further
representation of men and women. The study
subjects known to them, so the sample increases in
sample is then obtained by taking equal sample
size like a rolling snowball.
sizes from each stratum.
Non-Probability Sampling
Convenience Sampling
Convenience sampling is perhaps the
easiest method of sampling because participants
are selected based on availability and willingness to
take part.
Useful results can be obtained, but the
results are prone to significant bias, because those
who volunteer to take part may be different from
those who choose not to (volunteer bias), and the
sample may not be representative of other
characteristics, such as age or sex.
NOTE: Volunteer bias is a risk of all non-probability
sampling methods.
Quota Sampling
This method of sampling is often used by
market researchers. Interviewers are given a quota
of subjects of a specified type to attempt to recruit.
While this has the advantage of being
relatively straightforward and potentially
representative, the chosen sample may not be 1. Probability Sampling
Probability sampling refers to techniques where each member ordered by age or location, the sample may
of the population has a known, non-zero chance of being not be representative.
selected. These methods are often used when the goal is to
make generalizations about the population based on the c. Stratified Sampling
sample. The main advantage is that they are unbiased, and
• Definition: In stratified sampling, the population is
statistical inference methods can be applied.
divided into subgroups (or strata) based on a specific
a. Simple Random Sampling characteristic (e.g., age, gender, income), and then a
random sample is selected from each subgroup.
• Definition: This is the most basic form of probability
sampling. Every member of the population has an • How it works:
equal chance of being selected.
o Divide the population into distinct strata (e.g.,
• How it works: If you have a population of 1,000 young vs. old, male vs. female).
people, you would assign each person a number (1- o Then, randomly sample from each stratum,
1,000) and then randomly select numbers (either ensuring that every subgroup is represented.
using a random number generator or drawing lots).
o The number of people selected from each
• Advantages: group could be proportional to the group
size, or it could be equal (if you want to
o Fair and unbiased because everyone has the
compare groups evenly).
same chance.
o The results can be generalized to the entire • Advantages:
population. o Ensures that all significant subgroups are
included, making the sample more
• Limitations:
representative.
o If the population is large, it can be impractical
o Useful for comparing specific subgroups.
or costly to list every member and then
select from that list. • Limitations:
b. Systematic Sampling o Requires detailed information about the
population to divide it into appropriate strata.
• Definition: In systematic sampling, you select every
kkk-th individual from a list, after choosing a random o Can be more complex and time-consuming
starting point. than other methods.
• How it works: d. Cluster Sampling
o First, decide on the sampling interval kkk, • Definition: In cluster sampling, the population is
where kkk is the ratio of the population size divided into clusters (often geographically or naturally
to the sample size. For instance, if the occurring groups), and a random sample of clusters is
population size is 500 and you want a selected. Then, all or a random sample of individuals
sample of 50, you choose every 10th person within the selected clusters is surveyed.
(k=10k = 10k=10).
• How it works:
o Then, randomly select a starting point
between 1 and kkk, say person number 3, o Divide the population into clusters (e.g.,
and then choose every 10th person after that schools, neighborhoods, cities).
(3, 13, 23, 33, etc.).
o Randomly select a few clusters.
• Advantages:
o Survey all or a random sample from the
o Easier than simple random sampling, selected clusters.
especially if there’s a large list of people.
• Advantages:
o Ensures that the sample is spread
throughout the population. o More efficient and cost-effective than other
sampling methods, especially when the
• Limitations: population is spread out geographically.
o Can introduce bias if there’s a hidden pattern o Allows for surveying large populations with
in the population. For example, if the list has fewer resources.
a regular pattern, such as people who are
• Limitations: o Very quick and cost-effective.
o Less precise than stratified sampling, as the o Useful for exploratory research or when the
individuals within clusters may be more goal is not to generalize to the larger
similar to each other than to the population at population.
large.
• Limitations:
o Potential for increased sampling error.
o Highly biased and not representative of the
e. Multistage Sampling population.
• Definition: Multistage sampling is a combination of o Results may not be generalizable.
multiple sampling methods. It’s often used when a
b. Judgmental (Purposive) Sampling
population is geographically dispersed, and a more
complex sampling structure is needed. • Definition: The researcher deliberately selects
individuals who possess specific characteristics or
• How it works:
knowledge related to the research.
o First, use cluster sampling to randomly select
larger clusters (e.g., states or regions). • How it works:
o Then, within those clusters, use another o The researcher uses their expertise or
method like simple random sampling or judgment to choose a sample that they
stratified sampling to select smaller clusters believe will provide valuable insights.
or individuals. o For example, selecting expert participants in
a focus group about a new technology.
• Advantages:
o Very flexible and can be adapted to various • Advantages:
populations. o Focuses on gaining insights from specific
o Helps reduce cost and time, especially for types of individuals.
large populations. o Useful when targeting specific expertise or
experiences.
• Limitations:
o The complexity of multiple stages may • Limitations:
introduce errors. o Highly subjective and can be biased.
o There’s a risk of selecting less diverse or o Not generalizable because the sample is not
representative samples at each stage. randomly selected.
c. Quota Sampling
2. Non-Probability Sampling
• Definition: The researcher selects participants based
Non-probability sampling methods do not give every individual on specific characteristics until a predetermined quota
a known or equal chance of being selected. These methods for each characteristic is met.
are typically used when it's not feasible or necessary to
conduct a probability sample, or when the researcher is
• How it works:
focused on specific individuals rather than representing the o The researcher divides the population into
whole population. groups based on certain traits (e.g., age,
gender).
a. Convenience Sampling
o They then select individuals non-randomly
• Definition: The researcher selects individuals based
until they fill the desired number of
on convenience or accessibility rather than a
participants for each group.
structured method.
• Advantages:
• How it works:
o Ensures certain subgroups are represented.
o You choose people who are easiest to
access—people who happen to be nearby, o Quicker and more cost-effective than random
willing to participate, or easy to recruit. sampling.
• Advantages:
• Limitations: o Highly biased because people who volunteer
may have strong opinions or characteristics
o Not random, so it can still introduce bias. that differ from those who don’t volunteer.
o May not represent the full diversity of the o The sample may not represent the larger
population within the strata. population.
d. Snowball Sampling
• Definition: Snowball sampling is used when the Summary of Differences:
population is hard to reach or hidden. The researcher
starts with one or a few participants and asks them to • Probability Sampling methods are typically more
refer others to participate. reliable and statistically valid, ensuring that results
can be generalized to the broader population. They
• How it works: include simple random, systematic, stratified,
cluster, and multistage sampling.
o Start with an initial participant who meets the
criteria. • Non-Probability Sampling methods are less reliable
for generalization because not every member of the
o That participant then refers more individuals,
population has a known chance of being selected.
and this process continues like a snowball,
These include convenience, judgmental, quota,
with the sample growing over time.
snowball, and self-selection sampling.
• Advantages:
Each method has its place depending on the research
o Particularly useful for studying hidden or objectives, available resources, and desired accuracy of the
hard-to-reach populations (e.g., people with results.
rare diseases or members of underground
groups).
o Can be efficient when dealing with small or
hard-to-find populations.
• Limitations:
o Can introduce bias if the initial participants
are not representative.
o Difficult to control the sample size or
composition.
e. Self-Selection Sampling
• Definition: Participants volunteer to take part in the
study, typically through an open invitation (e.g.,
surveys posted online).
• How it works:
o You provide an opportunity for people to join
the study, often through an ad or call for
participants.
o Participants choose themselves to be part of
the sample.
• Advantages:
o Easy to implement, particularly for online
surveys.
o Cost-effective and quick.
• Limitations: