6 Sources of Experimental Error
When performing an experiment nothing can be measured with perfect accuracy. Experiments have to be designed with care to minimize potential sources of error that could skew the results away from the “true” value. Possible reasons for unexpected results might be:
- Human error (e.g. errors when following the procedure, errors in preparing solutions).
- Small sample size or differences due to biological variation
- Results were too variable to draw clear conclusions
The following Tables 1 and 2 cover common types of experimental error and other factors that might impact experimental results and how to minimize errors with experimental design.
Table 1
Types of Experimental Errors and How to Minimize them with Experimental Design and Statistics
| Types |
Random Errors |
Systematic Error |
| Definition | Random Errors are caused by unknown or unpredictable conditions in the experiment. | Systematic Error is an error in the experimental design or set-up which skews the results in the same direction each time. |
| Examples | Fluctuations in the readings from an instrument (depends on the accuracy of the instrument) or differences due to biological variation (see info below). | Incorrectly calibrated instruments, timing error, incorrect measurements or failure to choose a representative sample. Note: see information on human error (table 2). |
| How to minimize? | Collect more data. | This type of error cannot be corrected for statistically because all values are skewed in the same direction (i.e. too high or too low). |
| What to do… | Random errors can be evaluated through statistical analysis and reduced by averaging with a large sample size. For our experiment, calculate the mean, standard deviation, variance and 95% C.I.’s of the sample mean. | Design your experiment carefully and watching out for inconsistencies in the set-up that could skew your results before you collect your data. If a systematic error is uncovered while collecting data note it as it may influence your results. Depending on the error you may need to redo part or all of your experiment with an improved protocol. |
Table 2
Other Types of Factors that Influence Experimental Design
|
Biological Variation |
Human Error |
|
| Definition | Biological variation is any difference between individual organisms in a population caused either by genetic differences or environmental factors. Differences may be reflected in physical appearance, metabolism, behavior, or other measurable characters. | Human Error occurs when an experimenter makes an error in carrying out a procedure or experiment. Can be systematic if the experimenter makes the same error for all measurements. |
| Example | Humans metabolize caffeine at different rates. | Experimenter adds the wrong concentration of a chemical to a sample. Systematic if they add the wrong concentration to all samples. |
| How to minimize? | Collect more data to capture the range in variation through statistical analysis and ensure that you are collecting a random sample of the study population. | Prepare for experiment ahead of time and carefully follow procedure. |
| What to do… | Depending on the question being asked select organisms that are uniform. For example, a in a study examining how humans metabolize caffeine, limit your study population to include only individuals with no known metabolic disorders. | Depending on the error you may need to redo part of your experiment. All errors in how you carry out a procedure should be noted as they may influence your results. |
NOTE
Think carefully before attributing unusual results to “human error”. There may be a more interesting scientific explanation for the results that you found. Make sure to record any error you make, as you must discuss in detail the impact this error had on your results using your scientific knowledge. You will also want to discuss how you would go about testing any explanations you give for unexpected results.