Professional Documents
Culture Documents
Venkatesh Shankar
Identify the type of scale (nominal, ordinal, interval, ratio) being used in each of the
following questions. Justify your answer.
1. The analysis for each of the questions is given below the corresponding question. Is
the analysis appropriate for the scale used?
About 50% of the sample was born in the fall, 25% of the sample was born in the
spring, and the remaining 25% was born in the winter. It can be concluded that the
fall is twice as popular as the spring and the summer seasons.
The average income is $25,000. There are twice as many households with an income
of less than $9,999 than households with an income of $40,000 and over.
c. Which are your three most preferred brands of cars? Rank them from 1 to 3
according to your preference, with 1 as most preferred,
____ Toyota ____Pontiac ____ Saturn
____ Honda ____Buick ____ Chrysler
____ Nissan ____Ford ____ Dodge
d. How much time do you spend a day traveling to school every day?
____under 5 min ____16-20 min
____5-10 min ____over 20 min
____11-15 min
The median time spent on traveling to school is 8.5 minutes. There are three times as
many respondents traveling less than 5 minutes than respondents traveling 16-20
minutes.
The average satisfaction score is 4.5, which seems to indicate a high level of
satisfaction with Newsweek magazine.
2
10% of the respondents drink less than 2 cans of colas a day whereas three times as
many respondents drink over 5 cans of cola a day.
Marketing research is the most frequently taken course because the median is 3.2.
The response indicates that 40% of the sample have some high school education, 25%
of the sample are high school graduates, 20% have some college education, and 10%
are college graduates. The mean education level is 2.6.
Suppose you see an advertisement that claims that Vital capsules are 50 per cent more
effective in easing tension than the leading tranquilizers, Restease. As the research
director of the firm that produces Restease, you immediately begin comparison tests.
Using large sample sizes and a well-designed experiment, you have one group of
individuals use Vital capsules and a second group use Restease. You then have each
individual in each group rate the effectiveness of the brand they tried on a five-point
scale as follows:
For analysis, you decide to code the “very effective” response as +2; the “effective”
response as +1; the “neither-nor” response as 0; the “ineffective” response as -1; and
the “very ineffective” response as -2. This is a common way of coding data of this
nature.
3
You calculate an average response for Vital and Restease and obtain scores of 1.2 and
0.8 respectively. Because the 0.4 difference is 50 per cent more than the 0.8 level
obtained by your brand, you conclude that the claims for Vital are valid. Shortly after
reaching this conclusion, one of your assistants, who was also analyzing they data,
enters your office with the good news that Vital was viewed as only 10.5 per cent
more effective than Restease. Immediately, you examine his figures. He used that
same data and made no computational mistakes. The only difference was that he
assigned a “very ineffective” response at +1 and continued up to a +5 for the “very
effective” response. This is also a widely used procedure.
Then, as you are puzzling over these results, another member of your department
enters. She used the same approach as your assistant but assigned a +5 to “very
ineffective” and a +1 to “very effective.” Again, with no computational errors, she
found Vital to be 18.2 per cent more effective. What do you conclude?
Source: Derived from B. Venkatesh, “Unthinking Data Interpretation Can Destroy Value of Research,”
Marketing News, January 27, 1978, 6, 9. Both brand names are completely fictitious.