Professional Documents
Culture Documents
Mk0004 Solved Assignment
Mk0004 Solved Assignment
Mk0004 Solved Assignment
Test of Hypotheses
Hypothesis testing determines the validity of the assumption (technically described as null
hypothesis) with a view to choosing between the conflicting hypotheses about the value of the
population hypotheses and about the value of the population parameter. Hypothesis testing helps
to determine on the basis of sample data, whether a hypothesis about the population is likely to
be true or false. Statisticians have developed several tests of hypotheses (also known as tests of
significance) which can be classified as:
a) Parametric tests or standard tests of hypotheses;
b) Non parametric tests or distribution-free tests of the hypotheses.
Parametric tests usually assume certain properties of the parent population from which we draw
samples. Assumptions like observations come from a normal population, sample size is large,
assumptions about the population parameters like mean, variants, etc. must hold good before
parametric tests can be used. However, there are situations when the researcher cannot or does
not want to make assumptions. In such situations we use statistical methods for testing
hypotheses which are called non-parametric tests, because such tests do not depend on any
assumption about the parameters of parent population. Besides, most non-parametric tests
assume only nominal or original data, whereas parametric tests require measurement equivalent
to at least an interval scale. As a result, a non-parametric test needs more observation than a
parametric test to achieve the same size of Type I and Type II error.
Important Parametric Tests
The important parametric tests are:
1) z-test
2) t-test
3) x2-test
4) f-test
All these tests are based on the assumption of normality, i.e., the source of data is considered to
be normally distributed. In some cases the population may not be normally distributed, yet the
test will be applicable on account of the fact that we mostly deal with samples and the sampling
distributions closely approach normal distributions.
Ques.3 Discuss the different types of direct response attitude scales with examples?
Ans.3 Direct Response Attitude Scales
There are many ways to present a respondent with a continuum of numbered categories that
represent the range of possible attitude judgments. They can be generally classified as single item
scales and multiple item scales.
1) Single Item Scales
Single item scales are those that have only one item to measure a construct. Under the single
item scales, the itemized category scale is the most widely used by marketing researchers. In
some situations, comparative scales, rank order scales, or constant-sum scales have advantages.
2) Itemized Scales
There are four categories from which respondents can choose to indicate their overall level of
satisfaction with their present health insurance plan.
– Very satisfied
– Quite satisfied
– Somewhat satisfied
– Not at all satisfied.
This satisfaction scale has the following characteristics -
· All categories are labeled.
· The respondent is forced to make a choice; there is no provision for neutral opinion or ‘don’t
know’ response.
· There are more favorable than unfavorable categories, so the scale is unbalanced.
· There is no explicit comparison of respondents’ present plan with other health insurance plans.
3) Comparative Scales
Another version of the preceding scale would label the categories “excellent”, “very good”,
“fair” and “poor”, thereby eliminating the implicit comparisons. The problem with a comparative
scale is that the reference point is unclear and different respondents may use different reference
points or standards.
4) Rank Order Scales
Rank order scales require the respondent to arrange a set of objects with regard to a common
criterion: advertisements in terms of interest, product features in terms of importance, or new-
product concepts with regard to willingness to buy in the future. The result is an ordinal scale
with the inherent limitations of weak scale properties. Ranking is widely used in surveys,
however, because it corresponds to the choice process occurring in a shopping environment,
where a buyer makes a direct comparison among competing alternatives (brands, flavors, product
variations, and so on).
5) Q-Sort Scaling
When the number of objects or characteristics that are to be rated or ranked is very large, it
becomes rather tedious for the respondents to rank order or to do pair-wise comparisons. If the
respondent is forced to do rank ordering or a pair-wise comparison, a number of problems and
biases creep into the study. To deal with such situations, the Q-sort scaling process is used. In
Q-sort scaling, the respondents are asked to sort the various characteristics or objects that are
being compared into various groups, such that the distribution of the number of objects or
characteristics in each group follows a normal distribution. For example, let us take the case of a
toy manufacturing company such as Toys ‘R us’ developing a new product. After a marathon
brain storming session, the new product team has come up with a hundred different products,
each with minor variations in features, and wants to test and find out from consumers which
feature combination is most preferred and will generate maximum sales. The best scaling
procedure that can be used in this context is Q-sort scaling.
6) Constant Sum Scales
Constant sum scales require respondents to allocate a fixed number of rating points (usually
hundred) among several objects, to reflect the relative preference of each object. It is widely used
to measure the relative importance of attributes.
7) Pictorial Scales
In the pictorial scales, the various categories of the scale are depicted pictorially. The
respondents are shown a concept, or read an attitudinal statement and are asked to indicate their
degree of agreement or interest, by indicating the corresponding position on the pictorial scale.
8) Paired Comparison Scales
The brands to be rated are presented two at a time, so each brand in the category is compared
once to every other brand. In each pair, the respondent is asked to divide ten points among the
brands on the basis of how much they like one compared to the other. A score is then totaled for
each brand. Although, this scale performs well on the criteria, it is cumbersome to administrate.
Another possible limitation is that the frame of reference is always the other brands in the set
being tested. These brands may change over time.
9) Derived Attitude Scales
Attitudes towards complex objects such as health plans, automobiles, credit instruments, or
transportation modes have many facets. Thus, it is often unrealistic to attempt to capture the full
picture with one overall attitude-scale question. For example, the public appears to support the
general idea of income tax reforms, but opposes the elimination of the most popular tax
loopholes. The most frequently employed of these methods are the Likert, Thurstone and
semantic-differential scales.