Satisfactory, Unsatisfactory) are an interval scale.
2. Satisfaction measured using Bipolar Likert (Very
Satisfactory, Satisfactory.etc) is an interval scale.
3. A semantic differential like the one
below serves like an interval scale though strictly it may not be I find the MBA program @ IFMR (please tick only one of the 5 boxes below). Very much Interesti ng
Somewh at
Neither
Somewh
Very
at
much Boring
4. Even if we add up the number of correct answers
in test, not all problems are equally difficult and thus that 80% right answers is not twice as good as 40% right answers. So test scores are definitely ratio scales.
5. If you want to use the normal distribution and
related tests, the data must have mean and standard deviation. So you cannot use data from nominal and ordinal scales if you want to use normal distribution, correlation, regression etc.
6. I make a test of 30 items all about multiplying (say 22 x 15 etc.).
Then using random numbers I rearrange them in 10 different ways, so that there is no sequence effect (I am worried that students may get tired as the test progresses). Then I administer the 10 tests to 10 students each i.e. to 100 students. I find out the proportion of students who correctly answer each question. On that basis I look for questions that have been answered correctly by the same proportion of students and classify them as being as the same level of difficulty. Now I found 10 questions of the same level of difficulty (i.e. 50% of the students answered each correctly). Next I use these 10 questions to evaluate multiplication skills among students (not the same students but similar students). Thus I have created a ratio scale.
7. In the interval scale there is no absolute Zero.
8. I can use the Median only for Ordinal data. I cannot use the Median for Interval and Ratio data. 9. It is easy to find the standard deviation for both Nominal and Ordinal data. 10.The Pie Chart is best suited to deal with interval and ratio scale data. 11.The only measure of central tendency for the nominal scale data is the Median.
12.The higher level scales automatically have the benefits
of the lower level scales, for e.g. even in the ratio scale, the higher number data point is above the lower number data point. 13.The Median is always less than Mean 14.The Median is always somewhere between the Mean and the Mode 15.A data set can have only one Mode 16.The second quartile is more commonly known as the Median
17.Outliers are defined as data points that are greater
than the 3rd quartile or less than the 1 st quartile 18.In common language, when people say average, they are talking about the arithmetic mean 19.If we are looking at the short term yields in the stock market, we can recognize the arithmetic mean and the standard deviation, but if we are looking that the long term yield we must find the CAGR 20.An Ogive is constructed using the frequency distribution and a histogram is constructed using the cumulative frequency distribution