Question:medium

Ram, an economist, and Ramesh, an astrologer, had a debate. Ram said, "Astrology does not work. It just cannot predict." "It can predict better than your subject," rebutted Ramesh.
The evidence that best resolves the above debate will be:

Show Hint

To compare predictive accuracy fairly between two groups, always compare a rate or percentage, not a raw count.
Updated On: Jul 10, 2026
  • Conduct a survey among scientists asking which one of the two should be considered a science.
  • Compare the past performance of astrologers and economists in terms of the number of predictions which have come true.
  • Conduct a survey among economists asking their opinion regarding the ability of economic theory to predict economic phenomena.
  • Conduct an experiment where both astrologers and economists would be asked to predict the future. Compare the percentage of predictions that come true.
Show Solution

The Correct Option is D

Solution and Explanation

Think of this as designing a fair experiment. Ram and Ramesh disagree about accuracy, so whatever evidence we collect must isolate accuracy and nothing else, like sample size differences or personal opinions.

  1. Option A: A vote among scientists on which field deserves the label "science" is a question about status, not a measurement of predictive success. It settles nothing about accuracy.
  2. Option B: Counting raw correct predictions rewards whichever group predicts more often, even if their hit rate is lower. This is a flawed comparison because it is not adjusted for how many attempts each side made.
  3. Option C: This only surveys economists' self-opinion, ignoring astrologers entirely and relying on belief rather than actual outcomes.
  4. Option D (officially option E): Have both sides make predictions under the same conditions, then compare what percentage of each side's predictions actually came true. This controls for the number of attempts and gives a like-for-like accuracy comparison.

Because this option adjusts for sample size by using percentages instead of raw counts, it is the cleanest way to test who really predicts better.

Let's summarize:

  • A fair accuracy test must control for how many predictions are made.
  • Percentage correct does this; a raw count does not.
  • Opinion surveys (A, C) do not measure actual predictive accuracy at all.

The answer is the percentage-comparison experiment.

Was this answer helpful?
0

Top Questions on Critical Reasoning

Questions Asked in XAT exam

Read the scenario and answer the question:
The upper hinge of a dataset is the median of all the values to the right of the median of the dataset in an ascending arrangement, while the lower hinge of the dataset is the median of all the values to the left of the median of the dataset in the same arrangement.
For example, consider the dataset 4, 3, 2, 6, 4, 2, 7. When arranged in the ascending order, it becomes 2, 2, 3, 4, 4, 6, 7. The median is 4 (the bold value), and hence the upper hinge is the median of 4, 6, 7, i.e., 6. Similarly, the lower hinge is 2. 
A student has surveyed thirteen of her teachers, and recorded their work experience (in integer years). Two of the values recorded by the student got smudged, and she cannot recall those values. All she remembers is that those two values were unequal, so let us write them as A and B, where A < B. The remaining eleven values, as recorded, are: 5, 6, 7, 8, 12, 16, 19, 21, 21, 27, 29. Moreover, the student also remembers the following summary measures, calculated based on all the thirteen values: 
Minimum: 2
Lower Hinge: 6.5 
Median: 12 
Upper Hinge: 21 
Maximum: 29
View Solution
Read the following scenario and answer the questions that follow.
A T20 cricket match consists of two teams playing twenty overs each, numbered 1 to 20. The runs scored in any over is a non-negative integer. The run rate at the end of any over is the average runs scored up to and including that over, i.e., the run rate at the end of the kth over is the average number of runs scored in overs numbered \(1, 2, …, k,\) where \(1 \leq k \leq 20, k\) a positive integer.
The following table indicates the run rate of a team at the end of some of the overs during a T20 cricket match (correct up to 2 decimal places), where \(1 \leq N – 2 < N + 6 \leq 20\), N a positive integer. It is also known that the team did not score less than 6 runs and more than 15 runs in any over.
Over NumberRun Rate
N-28.00
N7.43
N+28.11
N+48.45
N+68.08
View Solution