You must set up an experiment to test your hypothesis.

Define the Variables

There should be three categories of variables in every experiment: dependent, independent, and controlled.

Dependent -- is what will be measured; it's what the investigator thinks will be affected during the experiment.

For example, the investigator may want to study coffee bean growth. Possible dependent variables include: number of beans, weight of the plant, leaf surface area, time to maturation, height of stem.

Independent -- is what is varied during the experiment; it is what the investigator thinks will affect the dependent variable.

In our coffee bean example, possible independent variables include: amount of fertilizer, type of fertilizer, temperature, amount of H,O, day length, all of these may affect the number of beans, weight of the plant, leaf area, etc.

Key : Since you need to know which factor is affecting the dependent variable(s), there may be only one independent variable. The investigator must choose the one that he/she thinks is most important. But the scientist can measure as many dependent variables as he/she thinks are important indicators of coffee bean growth.

Controlled -- the variables held constant. Since the investigator wants to study the effect of one particular independent variable, the possibility that other factors are affecting the outcome must be eliminated.

For example, the above scientist must ascertain that no differences in the type of fertilizer used exists, or amount of H,O, variations of temperature, or day length exist.

Outline Procedure

You must find a procedure or method to measure the dependent variable.

Procedures are developed by:

a. reading articles published by other scientists; b. talking to other scientists at universities/industries or at scientific meetings; c. one's own novel and creative ideas.

In the process of outlining a procedure, the following must also be determined:

a. the levels of treatment or the appropriate values to use for the independent variable; b. the numbers of replications or how many times the experiment will be repeated to ensure that the results are consistent; c. the control* treatments.

* A control is a "treatment" in which the independent variable is either eliminated or set at a standard value.

In the coffee bean experiment, if the researcher is hypothesizing that fertilizer affects bean yield, a possible control would be the treatment in which either no fertilizer is applied or some standard amount of fertilizer is applied.

This allows the scientist to ascertain that the effect on the dependent variable is in fact due to the independent variable.

Make Predictions

The scientist applies his/her present knowledge to predict the effect of the independent variable or the dependent variable.

The prediction is a statement of the expected results of the experiment based on the hypothesis. The prediction is often an "if/then statement."

For example: If increasing fertilizer increases number of beans, then coffee bean plants treated with more fertilizer will have more beans.

Predictions provide a reference point for the scientist. If predictions are confirmed, the scientist has supported the hypothesis. If the predictions are not supported, the hypothesis is falsified. Either way, the scientist has increased knowledge of the process being studied.

For example, according to the above prediction, you would expect this graph:

However, the actual data may produce this graph:

The scientist has learned that the prediction (greater applications of fertilizer caused increased number of beans) is true only up to a point. The scientist may now wish to identify this point specifically, i.e., find the optimum amount of fertilizer to apply. The scientist may also want to extend the research to a new direction and find out why higher fertilizer applications actually cease having an effect on the number of beans produced.

Write a Protocol

Once you have designed your experiment you need to formalIy present it in a protocol.

A protocol is simply a recipe, or written design, for performing the experiment.

You must write a protocol to insure that you have both a clear idea of how you will do the experiment and that you will have all the materials that are needed. A scientist usually writes his/her protocol in a laboratory notebook. Following the completion of the protocol, the next step in the scientific process is to perform the experiment. As the investigation takes place, observations are made and results are recorded.

Components of an Experimental Protocol

1. Purpose: This is a formal statement which encompasses your hypothesis. It is a statement of what question you are trying to answer and what hypothesis you wish to test.

2. Materials: List all major items needed to carry out your experiment. This list need not be lengthy if the materials are already published (as in the lab manual), but it should include the essentials.

3. Methods: How will you set up your experiment? How many experimental groups will you have? How will you measure the effect you wish to study? How long will the experiment last? These and any other methods should be explicitly stated or referenced from your lab manual so that a reader has all the information they need to know to be able to repeat your experiment and verify your results.

4. Controls: Identify the relevant control(s). Think about the variable(s) you and your group are manipulating. Your control needs to be held under natural, or unmanipulated conditions, not affected by the tested variable.

5. Data Interpretation: What will be done with the data once it is collected? Data must be organized and summarized so that the scientist himself, and other researchers can determine if the hypothesis has been supported or negated. Results are usually shown in tables and graphs (figures).

6. References: Any published works that you cite in your protocol should be listed in the reference section so that anyone reading your protocol can look that work up if they desire. The most common reference you will have is this manual.

Putting this all together, the scientist will be able to write a scientific paper once his/ her data is collected. For these laboratories it should be possible to write a good protocol in less than a page. A sample protocol format has been written for your reference. Remember do not write fluff. The reader of a protocol is interested in being informed concisely and accurately.

SAMPLE PROTOCOL:

Purpose:

We have made the observation that not all coffee bean plants mature identically. After conducting some background literature, we have come up with the following hypothesis: Nutrient resources in fertilizers are essential to coffee bean growth, lack of fertilizer retards growth.

Materials:

20 coffee bean seeds, soil from a constant source, fertilizer with a known amount of nitrogen and phosphorus, pots to plant the seeds, a constant UV light source.

Methods:

1.There will be two groups of seeds with 10 plants each: a) the seeds which have fertilizer (independent variable); and b) the seeds which do not have fertilizer (control group).

2. Plant all seeds in 30 cm diameter pots with soil. The fertilizer treatment will receive 10 grams of fertilizer.

3. At the end of the growing season, the number of beans(dependent variable)of each plant will be recorded.

Control:

The plants will have natural (no fertilizer) soil conditions.

Data Interpretation:

A histogram will be used to plot the results. The average number of beans for each group of plants will be plotted on the Y axis (ordinate) and the treatment group will be plotted on the X axis (abscissa). A t-test can also be performed to determine if the treatment group differs from the control group. If the treatment group produces more seeds than the control, we can then conclude that the treatment of fertilizer had an effect and the resource in question is limited to plants. This will obviously lead to more questions and hypotheses.

References:

Morgan, I.G. and Brown Carter, M.E., Investigating Biology: A LaboratoryManual for Biology. California: Benjamin/Cummings Publishing Co., Inc. 1993.

Evaluate the Data

A well-designed experiment will produce mounds of data. This data is like the forensic evidence at a murder trial, and your role is like both the prosecution and the defense. You must weigh the evidence. Evaluate the data to determine whether or not it supports your hypothesis. Sort through it all, and organize it into a meaningful presentation of tables and/or figures. Sometimes statistical tests or other calculations are required to evaluate the significance of the results.

Make Conclusions

If the data supports your hypothesis, you are on to something meaningful. If not, consider why, based on your knowledge of the system. Then it is back to the drawing board-time to make another hypothesis. Perhaps you cannot trust your results due to some questionable procedures, and you should repeat the experiment. Maybe you find you do not have enough information; collect more data from the same experiment, or design a new one to test the same hypothesis. Or maybe your results bring up many exciting new questions that require you to do scores of new experiments to answer them!!!

You should get the idea that whatever your conclusion is, it is never final. You should always end up with new questions, with some components yet unresolved. The more we know, the more we know how much more we do not know.

Return to Table of Contents