Skip to Main Content

Biology IA: Recording and Analyzing Data

Guide for the Internal Assessment Individual Project in IB Biology

Lab Design and Data Analysis

This page gives tips on planning your project:

  • Designing your lab to minimize errors
  • Maximizing precision and minimizing uncertainty when collecting data
  • Knowing when you've collected enough data
  • Conducting the appropriate analysis of your data

Lab Design: Reducing errors

Discuss in your procedures the steps you are taking in your experiment to reduce errors.  Address all of the following types of error as they apply to your experiment:

Random, normal variation

  • Living organisms exhibit random, normal variation.  Not every bean plant, for example, will respond to a manipulated variable in the same way.  Not every cell has the same tonicity.  Not every grass field has the same soil conditions.  How will you account for normal variation in your lab design?

Human error

  • Humans make mistakes.  How will you reduce human error in your lab design?

The effects of measurement

  • The act of measuring something may change the very variable being measured.  For example, a cold thermometer inserted into a substance may cool that substance.  How will your lab design reduce or monitor these effects?

Equipment error

  • How will you test your equipment for accuracy?  Does it require calibration?

Collection of Data: Precision & Uncertainty

In general, the precision of a tool is plus or minus half of the smallest division on the instrument.  If a thermometer reads in degrees, the precision for the thermometer is +/- 0.5 degree.  When recording a temperature, extend the significant digits to tenths of a degree to match this level of precision.   Here is an example:  14.0 ° +/- 0.5 °C

Since one must estimate the reading on a ruler at both ends of the object, the precision of a ruler is +/- the smallest increment on the ruler (2 times half the smallest increment).  Here is an example:  42 mm  +/- 1 mm.

Find the manufacturer’s estimate of precision for electronic instruments.

Be careful to be precise when measuring.  Read the bottom of a meniscus, for example.  Hold a thermometer in the substance being measured, not touching the glassware that holds the substance.  Take readings at eye level.

Estimate all sources of error in an overall estimate of uncertainty.  For example, a stop watch will have a precision based on the units given, but human reflex speed in starting and stopping the stopwatch will add additional uncertainty.

How Many Data Are Enough?

As a general rule of thumb, test a minimum of five variations in the independent variable and make at least three measurements each.  For example, measure a rate of reaction at a minimum of five temperatures, three times at each temperature.

Create a histogram/frequency distribution of your data.  In general, living systems will give data that fall in a rough approximation of a normal distribution (bell curve).  Significant variation from a normal distribution may indicate the need for more data collection.

Calculate standard deviation.  If the standard deviation is very large compared to the means of your measurements, this may indicate the need for more data collection.

In general, school laboratory time is limiting.  Collect as many data as you can and then push yourself to collect more.

Britannica Image Quest - Computer Data

Analyzing Data

When you want to see if the differences between data sets are statistically significant, conduct a t-test to find a p-value.

When you want to see if different sets of data are independent, perform a chi-squared test.  Here are some examples:  Are two species in a habitat distributed independently or do they tend to associate with each other?  Do two genetic traits assort independently or do are they linked?

When you want to see if one variable correlates with another, calculate a correlation coefficient, r.

Britannica ImageQuest - Safety signs seen on a laboratory door