How to Distinguish Between Data & the Interpretation of the Data

Data is quantitative information, while its interpretation is subjective.

The point of polls and experiments is to collect data. Data are hard information that is used to support an answer to a specific experimental question; nothing in science is ever proven definitively. For example, an experiment to test the effect of smoking on lung cancer would collect data consisting of whether a person smoked, and if he had lung cancer during his life. Those two items of information are the data. Everything else is an interpretation of that data.

Instructions

1. Determine if the item makes a moral statement. If an item refers to something being "right or wrong", or "good or bad", then that is an interpretive statement. It is not a piece of data. The only exception to this is if the data you are looking at is the answer to a poll question from an individual correspondent. If that is the case, then these statements are still data.

2. Determine if the item makes a qualitative statement about what the entire poll or entire study means, as opposed to a quantitative statement. A quantitative statement would be "More people chose option 'A' over option 'B'." This is a fact that can be born out by arithmetic. Statements like "People prefer option 'A' over option 'B'" are immediately suspect, since they are making assumptions about a respondent's state of mind.

3. Look for charged words. Data and statements relating it should be bland and unexciting. For example, "More respondents chose option 'A'" vs "Respondents overwhelmingly rejected option 'B'". The later statement interprets the thought process of respondents.

4. Look for quantitative information. Numbers do not lie or obfuscate, whereas words always have multiple possible meanings.

Tips & Warnings

• Always remember that data is un-opinionated and amoral. Any statement stating what data "means" or "suggests" is interpretation.

References

• University of Wisconson, Eau Claire: Data Interpretation Methods