How to Handle Unusual Data

From EG1003 Lab Manual
Revision as of 03:06, 11 July 2011 by Admin (talk | contribs) (5 revisions)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

How to Handle Unusual Data

Introduction

Sometimes when you're performing a lab, you get readings that don't appear to be right. This presents a dilemma. Many important discoveries are based on somebody noticing some data that doesn't seem to fit, leading to further investigation and the advancement of technology. On the other hand, frequently the data is just incorrect, giving misleading results. We need to be able to handle this situation in a way that doesn't detract from analyzing the results, but also doesn't destroy important evidence.

What to do

When you're performing a lab, ALL the data you record should go into your reports under the Data/Observations sections. Nothing should be omitted. That way, in case there's a question later, at least you've recorded the evidence.

After you've recorded all the data, you can then make a decision to omit data which is obviously wrong. For example, if you've immersed an item in hot water and then taken it out, the item cannot be hotter than the temperature of boiling water: 100°C or 212°F. If you record a temperature of 2000°F, that would be the temperature inside a steel blast furnace, and the object probably would have been destroyed! This data is obviously incorrect.

In your "Results" or "Discussion" section of your reports, you should make note of the erroneous reading and remove it from the data. You should then proceed with any other calculations that you need to make on the remaining data.

Summary

You've now avoided the dilemma discussed at the beginning of this page. You've recorded the data, so that if there's a question later the original data is still present. However, you've also avoided having what appears to be incorrect data cause you to make incorrect calculations or draw incorrect conclusions about what you've done.

Return to Table of Contents