UNIT 5

Measurement, Scaling &

Sampling
 Variables
 In simple term, A variable in research simply refers to a
person, place, thing, or phenomenon that you are trying to
measure in some way.
 A variable is defined as anything that has a quantity or
quality that varies.
 Values of the variable differs from one activity to next
activity.
 In short, a variable represents a measurable attribute that
changes or varies across the experiment whether comparing
results between multiple groups, multiple people or even
when using a single person in an experiment conducted over
time.
 Types Of Variables

 Dependent variable
 Independent variable
 Moderating variable
 Intervening variable
 Types Of Variables

 The dependent variable is the variable a researcher is

interested in. The changes to the dependent variable are what
the researcher is trying to measure with all their fancy
techniques.
 An independent variable is a variable believed to affect the
dependent variable. This is the variable that you, the
researcher, will manipulate to see if it makes the dependent
variable change.
 Types Of Variables

 A moderating variable affects the nature of the relationship

between independent and dependent variable.
 Moderator variables influence the relationship between the
independent and dependent variables. Researchers measure
moderator variables and take them into consideration during
the experiment.
 Types Of Variables
 Intervening variables link the independent and dependent
variables, but as abstract processes, they are not directly
observable during the experiment.
 Intervening variables explains a relation or provides a
causal link between other variables and moderating variable
is that influences, or moderates, the relation between two
other variables and thus produces an interaction effect.
 Types Of Variables

 For example, if studying the use of a specific teaching

technique for its effectiveness, the technique represents the
independent variable, while the completion of the technique's
objectives by the study participants represents the dependent
variable, while the actual processes used internally by the
students to learn the subject matter represents the intervening
variables
 Measurement

 Measurement is the process of observing and recording the

observations that are collected as part of research. The
recording of the observations may be in terms of numbers or
other symbols to characteristics of objects according to
certain prescribed rules. The respondent’s, characteristics are
feelings, attitudes, opinions etc.
 The most important aspect of measurement is the
specification of rules for assigning numbers to characteristics.
The rules for assigning numbers should be standardized and
applied uniformly. This must not change over time or objects.
 Characteristics of good Measurement

 Validity
 Reliability
 Practicability
 Scaling

 Scaling may be considered an extension of measurement.

 It is the assignment of objects to numbers or semantics
according to a rule. In scaling, the objects are text statements,
usually statements of attitude, opinion, or feeling.
 For example, consider a scale from 1 to 10 for locating
consumer preference for the product. Each respondent is
assigned a number from 1 to 10 denoting the degree of
unfavorableness for the product, with ‘1’ indicating extremely
unfavorable and ’10’ indicating extremely favorable. Here,
the measurement is the process of assigning the actual
number from 1 to 10 to each respondent while the scaling is a
process of placing respondents on a continuum with respect
to their preference for the product.
 Types of Scaling

 Nominal Scale is the crudest among all measurement scales

but it is also the simplest scale. In this scale the different
scores on a measurement simply indicate different categories.
 The nominal scale does not express any values or
relationships between variables.
 Types of Scaling

 The nominal scale is often referred to as a categorical scale.

The assigned numbers have no arithmetic properties and act
only as labels. The only statistical operation that can be
performed on nominal scales is a frequency count.
 Types of Scaling

 For example: labeling men as ‘1’ and women as ‘2’ which is

the most common way of labeling gender for data recording
purpose does not mean women are ‘twice something or other’
than men. Nor it suggests that men are somehow ‘better’ than
women.
 Types of Scaling
 Ordinal Scale involves the ranking of items along the
continuum of the characteristic being scaled. In this scale, the
items are classified according to whether they have more or
less of a characteristic.
 The main characteristic of the ordinal scale is that the
categories have a logical or ordered relationship. This type of
scale permits the measurement of degrees of difference, (i.e.
‘more’ or ‘less’) but not the specific amount of differences
(i.e. how much ‘more’ or ‘less’).
 Types of Scaling

 Using ordinal scale data, we can perform statistical analysis

like Median and Mode, but not the Mean.
 For example, a fast food home delivery shop may wish to ask
its customers: How would you rate the service of our staff?
 (1) Excellent • (2) Very Good • (3) Good • (4) Poor • (5)
Worst •
 Types of Scaling
 Interval Scale is a scale in which the numbers are used to
rank attributes such that numerically equal distances on the
scale represent equal distance in the characteristic being
measured.
 An interval scale contains all the information of an ordinal
scale, but it also allows to compare the difference/distance
between attributes.
 Types of Scaling

 The interval scales allow the calculation of averages like

Mean, Median and Mode and dispersion like Range and
Standard Deviation. For example, the difference between ‘1’
and ‘2’ is equal to the difference between ‘3’ and ‘4’.
 The only drawback of this scale is that there no pre-decided
starting point or a true zero value.
 Types of Scaling

 Temperature is a classic example of an interval scale because

we know that each degree is the same distance apart and we
can easily tell if one temperature is greater than, equal to, or
less than another. The difference between a temperature of
100 degrees and 90 degrees is the same difference as between
90 degrees and 80 degrees.
 Types of Scaling

 Temperature, however, has no absolute zero because there is

(theoretically) no point where temperature does not exist.
 Absolute zero refers to a point where none of the scale exists
or where a score of zero can be assigned.
 Types of Scaling
 A ratio scale possesses all the properties of the nominal,
ordinal and interval scales and, in addition, an absolute zero
point.
 It is the highest level of measurement scales. This has the
properties of an interval scale together with a fixed (absolute)
zero point. It is calculated by assuming that the variables
have an option for zero, the difference between the two
variables is the same and there is a specific order between the
options.
 Types of Scaling

 Ratio scales permit the researcher to compare both

differences in scores and relative magnitude of scores.
Examples of ratio scales include weights, lengths and times.
 In market research, a ratio scale is used to calculate market
share, annual sales, the price of an upcoming product, the
number of consumers, etc.
 Types of Scaling

 If you are 20 years old, you not only know that you are older
than someone who is 15 years old (magnitude) but you also
know that you are five years older (equal intervals). With a
ratio scale, we also have a point where none of the scale
exists; when a person is born his or her age is zero.
Difference between scales
 Attitude Measurement

 An attitude measurement survey is a study, on a properly

drawn sample, of a specified population to find out what
people in that population feel about a specified issue.
 Attitude surveys usually use carefully constructed,
standardized questionnaires.
 Te chnique for De veloping Attitude
Sc ale

 Choice
 Rating
 Ranking
 Single Choice

 What do you eat outside? (Tick just one)

 1. Breakfast
 2. Lunch
 3. Dinner
 4. In betweens
 Multiple choice

 While eating out which category of food do you generally

order? (You can tick as many as you like)
 1. Fast Food
 2. Italian
 3. Newari
 4. Thakali
 5. Indian
 6. Chinese
Rating
 Ranking

 Rank the companion below according to your preference

of eating out with them with (1-most preferred and 4-least
preferred)
 Family Members
 Friends
 Girlfriend/Boyfriend
 Co-workers/Staff
Construction of Attitude Scale
 Ar bitr ary Sc ale

 Arbitrary scales are developed on ad hoc basis and are

designed largely through the researcher’s own subjective
selection of items. The researcher first collects few
statements or items which he/she believes are unambiguous
and appropriate to a given topic.
 Some of these are selected for inclusion in the measuring
instrument and then people are asked to check in a list the
statements with which they agree.
 Arbitrary Scale

 The chief merit of such scales is that they can be developed

very easily, quickly and with relatively less expense.
Ar bitr ary Sc ale
 Differential Scales
 Under such an approach the selection of items is made by a
panel of judges who evaluate the items in terms of whether
they are relevant to the topic area and unambiguous in
implication.
 An important deterrent to their use is the cost and effort
required to develop them. Another weakness of such scales is
that the values assigned to various statements by the judges
may reflect their own attitudes.
 Differential Scales

 The researcher gathers a large number of statements, usually

twenty or more, that express various points of view toward a
group, institution, idea, or practice (i.e., statements belonging
to the topic area)
 Differential Scales
 These statements are then submitted to a panel of judges, each of whom
arranges them in eleven groups or piles ranging from one extreme to
another in position. Each of the judges is requested to place generally in
the first pile the statements which he thinks are most unfavourable to the
issue, in the second pile to place those statements which he thinks are
next most unfavourable and he goes on doing so in this manner till in the
eleventh pile he puts the statements which he considers to be the most
favourable.
 Differential Scales

 For items that are retained, each is given its median scale
value between one and eleven as established by the panel.
 A final selection of statements is then made. For this purpose
a sample of statements, whose median scores are spread
evenly from one extreme to the other is taken.
 Liker t Scale

 A Likert scale is what is termed a Summated instrument

scale. This means that the items making up a Liken scale are
summed to produce a total score. In fact, a Likert scale is a
composite of itemized scales. Typically, each scale item will
have 5 categories.
 Likert Scale
 The Likert scale requires the respondents to indicate a degree
of agreement or disagreement with each of a series of
statements about the stimulus objects.
 Cumulative Scale

 Essentially, we would like a set of items or statements so that

a respondent who agrees with any specific question in the list
will also agree with all previous questions.
 Put more formally, we would like to be able to predict item
responses perfectly knowing only the total score for the
respondent.
 Cumulative Scale

 The special feature of this type of scale is that statements in it

form a cumulative series. This, in other words, means that the
statements are related to one another in such a way that an
individual, who replies favourably to say item No. 3, also
replies favourably to items No. 2 and 1, and one who replies
favourably to item No. 4 also replies favourably to items No.
3, 2 and 1, and so on.
 Cumulative Scale

 For example, imagine a ten-item cumulative scale. If the

respondent scores a four, it should mean that he/she agreed
with the first four statements. If the respondent scores an
eight, it should mean they agreed with the first eight.
 The object is to find a set of items that perfectly matches this
pattern.
 Cumulative Scale

 The scale is designed to measure one factor or subject. For

example, the following shows a questionnaire for a person’s
attitudes towards depression:
 Semantic differential Scale

 The semantic differential is a seven-point rating scale with

end points associated with bipolar labels that have semantic
meaning.
Powerful --:--:--:--:-X-:--:--: Weak
Unreliable --:--:--:--:--:-X-:--: Reliable
Modern --:--:--:--:--:--:-X-: Old-fashioned
 Semantic differential Scale
 This scale is based on the presumption that an object can
have different dimensions of connotative meanings which can
be located in multidimensional property space, or what can be
called the semantic space in the context of S.D. scale
 This scaling consists of a set of bipolar rating scales, usually
of 7 points, by which one or more respondents rate one or
more concepts on each scale item.
 Semantic differential Scale

 The negative adjective or phrase sometimes appears at the

left side of the scale and sometimes at the right.
 This controls the tendency of some respondents, particularly
those with very positive or very negative attitudes, to mark
the right-hand or left-hand sides without reading the labels.
 Individual items on a semantic differential scale may be
scored on either a -3 to +3 or a 1 to 7 scale.
 Graphic Rating Scale

 A graphic rating scale is a commonly used scale system for

performance appraisals. The scale typically features a Likert
scale from 1-3, 1-5, and so on. An example of a 1-3 rating
could include responses such as: 1: Poor, 2: Average, and 3:
Excellent.
 Another example:
Poor Good
 Graphic Rating Scale

 This line can be vertical or horizontal and scale points may be

provided. No other indication is there on the continuous
scale.
 Its limitation is that coding and analysis will require
substantial amount of time, since we first have to measure the
physical distances on the scale for each respondent.
 What is Sampling?
 A sample is a subset of a population
 Statistical method of obtaining representative data or
observations from a group (lot, batch, population, or
universe)
 Shortcut method for investigating a whole population
 Example: We check a sample of rice to see whether the rice
well boiled or not.
 Why Sampling ?

 Less costs

 Less field time, energy

 More accuracy i.e. Can Do A

Better Job of Data Collection
 When it’s impossible to study the
whole population
 Sampling Types / Techniques
1. PROBABILITY-BASED ( RANDOM ) SAMPLING
1.1. Simple Random Sampling
1.2. Systematic Random Sampling
1.3. Stratified Random Sampling
1.4. Cluster Sampling
1.5. Multistage Sampling

2. NON-PROBABILITY ( NON-RANDOM )
SAMPLING
2.1. Purposive Or Judgmental Sampling
2.2. Convenience Sampling Or Accidental Sampling
2.3. Snowball Sampling
2.4. Quota Sampling
 1. PROBABILITY-BASED
( RANDOM ) SAMPLING
 These samples are based on probability theory
 Individual judgment plays no part in selection of sample
 Each item in the sample stands equal chance of being included in
the sample
 Reliable results of statistical analysis with measurable margins of
errors and degree of confidence
 Example: Long practiced various forms of random selection, such
as picking a name out of a hat, or choosing the short straw
 1.1. Simple Random Sampling
 Each unit in the population has an equal chance of being in
the sample

 The selection of each unit is independent of the selection of

every other unit

 Simple, easy, fair and reasonable way to select a sample

 Not the most statistically efficient method of sampling

 Just the luck of the draw, is not a good representation of

subgroups in a population.
 1.1. Simple Random Sampling
 Example: Selecting sample of 25 out
of 250 people

− Take 250 Ping-Pong balls

− Number then from 1 to 250
− Put them into a large barrel and mix
them up
− Grab 25 balls.
− Read off the numbers. Those are the
25 people in your sample.
− This is called the lottery method.
 1.2. Systematic Random Sampling
 Useful for selecting large samples, say 100 or more
 Every kth element in the frame is selected, where k, the

sampling interval (skip): is calculated as: k=N/n

 Much more efficient than simple random sampling
 Applied only on logically homogeneous population
 If the selection interval matches some pattern in the list
problem may arise resulting to biasness
 1.2. Systematic Random Sampling

 Example: Systematic
sample of people working
in an office

− Total number of people

working (N) = 50,
Sample size (n) = 5
− Every kth (50/5=10th)
member is selected
− Number 10, 20, 30, 40,
50 are selected from the
list of all population
members.
 1.3.Stratified Random
Sampling
 Involves dividing your population into homogeneous
subgroups
 Strata (two or more groups): according to some criterion,
such as geographic location, grade level, age, or income
 Subsamples are randomly selected from each strata.
 Used when the researcher wants to highlight specific
subgroups within the population.
 The population will be evenly sampled
 1.3. Stratified Random
Sampling
 Example: Stratified sample of clients
− The population of clients for our agency can be divided into
three groups: Caucasian, African-American and Hispanic-
American.
− Random sample is taken from each group/strata
 1.4. Cluster Sampling
 In statistics, cluster sampling is a sampling method in which
the entire population of the study is divided into externally
homogeneous but internally heterogeneous groups called
clusters.
 Essentially, each cluster is a mini-representation of the
entire population.
 1.4. Cluster Sampling

 It is often used to study large populations, particularly those

that are widely geographically dispersed. Researchers usually
use pre-existing units such as schools or cities as their
clusters.
 One or more clusters are chosen at random and everyone
within the chosen cluster is sampled.
 Reduce costs by increasing sampling efficiency
 1.4. Cluster Sampling
 After identifying the clusters, certain
clusters are chosen using simple
random sampling while the others
remain unrepresented in a study. After
the selection of the clusters, a
researcher must choose the appropriate
method to sample the elements from
each selected group.

 1.4. Cluster Sampling

 For example, if a researcher wants to conduct a study to judge

the performance of sophomore’s in business education across
Nepal, it is impossible to conduct a research study that
involves a sophomore in every college in Nepal. Instead, by
using cluster sampling, the researcher can club the colleges
from each city into one cluster. These clusters then define all
the sophomore student population.
 1.4. Cluster Sampling

 Let’s consider a scenario where an organization is looking to

survey the performance of smartphones across Germany.
They can divide the entire country’s population into cities
(clusters) and further select cities with the highest population
and also filter those using mobile devices.
Difference Between Cluster Sampling
and Stratified Sampling

 For a stratified random sample, a population is divided into

stratum, or sub-populations, before sampling. At first glance,
the two techniques seem very similar. However, in cluster
sampling the actual cluster is the sampling unit; in stratified
sampling, analysis is done on elements within each strata.
In cluster sampling, a researcher will only study selected
clusters; with stratified sampling, a random sample is drawn
from each strata.
 1.5. Multistage Sampling
 Multistage sampling divides large populations into stages to
make the sampling process more practical. A combination
of stratified sampling or cluster sampling and simple random
sampling is usually used.
 Stage in multistage Sampling
1. Clusters are formed.
2. In order to classify multistage sampling as probability
sampling, each stage must involve a probability sampling
method.
 1.5. Multistage Sampling
 While in the multistage sampling technique, the first level is
similar to that of the cluster sampling, where the clusters are
formed out of the population, but further, these clusters are
sub-divided into smaller targeting groups, i.e. sub-clusters
and then the subject from each sub-clusters are chosen
randomly.
 Further, the stages can be added depending on the nature of
research and the size of the population under study.
 1.5. Multistage Sampling
 For example, If the government wants to take a sample of
10,000 households residing in Gujarat state. At the first stage,
the state can be divided into the number districts, and then
few districts can be selected randomly. At the second-stage,
the chosen districts can be further sub-divided into the
number of villages and then the sample of few villages can be
taken at random. Now at the third-stage, the desired number
of households can be selected from the villages chosen at the
second stage.
 Your research objective is to evaluate online spending
patterns of households in the Nepal through online
questionnaires.
 NON-PROBABILITY (NON-
RANDOM) SAMPLING
 Does not give all the individuals in the population equal
chances of being selected
 Selection of elements are based on assumptions which forms
the criteria for selection
 Focus on volunteers, easily available units, or those present
when the research is done
 Useful for quick, inexpensive and convenient studies
 Example: If a survey is sent and then asked to pass the
survey onto a friend. It is biased and sample not defined.
 2.1. Purposive / Judgmental Sampling
 Selection is based on the knowledge of a population and the
purpose of the study.

 Usually used when a limited number of individuals possess

the trait of interest.
 Used for quick sampling
 Proportionality is not the primary concern
 Likely to get the opinions of your target population
 misrepresentation of the entire population
 2.1. Purposive / Judgmental Sampling

 Example:

− People in a mall or on the

street who are carrying a
clipboard and who are
stopping various people
and asking if they could
interview them?
− Looking for a targeted
group.
 2.2. Convenience/ Accidental Sampling

 Also called an “grab” or “opportunity” or "man-in-the-street"

samples
 Selection of whichever individuals are easiest to reach
 It is done at the “convenience” of the researcher
 No evidence that they are representative of the populations
 a sample population selected because it is readily available
and convenient
 2.2. Convenience/ Accidental Sampling
 Example: if the interviewer was to conduct such a survey
early in the morning on a given day
− the people interviewed would be limited to those present at
that given time
− would not represent the views of other people of that area
 2.3. Snowball Sampling
 Used for sample difficult to locate (hidden populations)
 But sometimes the best or only method available
 Hardly representative of the larger study population
 Existing study subjects recruit future subjects from among
their acquaintances
 Thus the sample group appears to grow like a rolling
snowball
 Subject to numerous biases like people with many friends are
more likely to be sampled
 2.3. Snowball Sampling
 Example: The book or the
movie “The Help”
− Skeeter (a white woman)
identified Aibileen (a
housekeeper), willing to speak
about her experiences working for
white families
− Aibileen, then recruited more
housekeepers for Skeeter to
interview
− Those housekeepers then recruited
a few more, and so on
 2.4. Quota Sampling
 Population is first segmented into mutually exclusive sub-set
 Then either convenience or judgment sampling is used to
choose people from each subset.
 The researcher decides how many of each category is
selected.
 Useful when:− Limited time with very tight research budget
− Detailed accuracy is not important.
 It is the non probability version of stratified sampling
 Each potential subject have an unknown probability of being
selected.
 2.4. Quota Sampling
 Example: An interviewer may be
told to sample 20 females and 30
males between the age of 45 and
60.
− Form a sub group of male and
female between the age of 45
and 60 in a locality
− Select 20 females from female
subset and 30 male from male
subset.
− This means that individuals
can put a demand on who they
want to sample on personal
preference
 Example

 Let’s assume your research objective is to evaluate the impact

of cross-cultural differences on employee motivation in
Virgin Media in the UK. You need to assess the effectiveness
of employee motivational tools taking into account gender
differences among the workforce.
1. Dividing the population into specific groups. Virgin Media
employees in the UK as the sampling frame need to be
divided into the following five groups according to their
cultural background:
 European
 Asian (India)
 Asian (China)
 Black (African)
 Other
2. Calculating a quota for each group.
 You supervisor confirms that in order to achieve research
objectives, 30 representatives from each group and the total
sample size of 150 respondents would be appropriate.
3. Determining specific condition(s) to be met and quota in
each group
 Both genders, males and females need to be represented
equally in your sample group. This is an important condition
that has to be satisfied. Accordingly, you recruit 15 males and
15 females from each group.
 Sampling Error

 The error that arises as a result of a sample taken from a

population rather then using the whole population.
− Population specification Error

− Sample frame error

− Selection error

− Non response

− Error in taking sample

 Population specification Error

 This error occurs when the researcher does not understand

who they should survey. For example, imagine a survey about
breakfast cereal consumption. Who to survey? It might be the
entire family, the mother, or the children. The mother might
make the purchase decision, but the children influence her
choice.
 Sample frame error

 A frame error occurs when the wrong sub-population is used

to select a sample. Frame error results when the sampling
frame is not an accurate and complete representation of the
population of interest.
 Frame errors arise from erroneous inclusions, erroneous
exclusions, and multiple inclusions.
 Sample frame error

 Erroneous inclusions occur if the frame includes units not in

the population of interest so that invalid units have a chance
of being included in the sample. Erroneous exclusions occur
when members of the population of interest are missing from
the list so that they have no chance of being sampled.
Multiple inclusions occur if some population members appear
multiple times on the list so that they have a higher chance of
being sampled.
 Example: College Alumni Study (Telephone Directory)
 Selection error

 This occurs when respondents self-select their participation in

the study – only those that are interested respond. Selection
error can be controlled by going extra lengths to get
participation. A typical survey process includes initiating pre-
survey contact requesting cooperation, actual surveying, and
post-survey follow-up. If a response is not received, a second
survey request follows, and perhaps interviews using
alternate modes such as telephone or person-to-person.
 Non response

 Non-response errors occur when the survey fails to get a

response to one, or possibly all, of the questions.. This may
occur because either the potential respondent was not
contacted or they refused to respond. The extent of this non-
response error can be checked through follow-up surveys
using alternate modes.
 Error in taking sample

 These errors occur because of variation in the number or

representativeness of the sample that responds. Sampling
errors can be controlled by (1) careful sample designs, (2)
large samples, and (3) multiple contacts to assure
representative response.
Me thods of Minimiz ing Sampling Error

Increase Unbiased
Cross Check
sample size Sampling

Appropriate
Clear
sampling
questionnaire
design
 Non Sampling Error

 Error which are incurred from other sources than selection of

sample
− Error of poor sampling design

− Over and under coverage

− Misinterpretation of Questions

− Processing errors

− Respondents related errors

− Measuring errors
Me thods of Minimiz ing Non- Sampling
Er ror
Checking the sampling process

Preparation of questionnaire

Pilot survey

Use of competent manpower

Provide information

Use of experts

Checking data processing and analyzing

Fix procedure