“Experimental design” refers to a method of planning an experiment, incorporating specific components. In Prince William County, we use the basic components of experimental design set forth by Julia Cothron, Ron Giese, and Dick Rezba in Students and Research (Kendall/Hunt, 1989).
The experimental design diagram is a good tool for students to use to plan an experiment. It is not intended to replace a lab report. The diagram reminds students to consider all aspects of the plan as they write it in a simple format. For the teacher, it provides a quick way to check whether the student has planned well to test only one variable at a time and to collect valid data.
The standard experimental design diagram is set up in a logical order to show the whole plan at a glance, but it is important to teach students that it is much easier to complete the parts in a different sequence. After the specific assignment or the Four-Question Strategy has led the students to define a problem to test, the first part of the diagram to be completed is the independent variable, followed by the independent variable. At this point students should specify the constants. A complete list of well-described constants is the key to determining the procedure for a well-designed experiment. Students find that it is much easier to write the title and the hypothesis after the other parts are decided.
Below is a copy of an experimental design diagram that enables one to visualize the components of the proposed experiment.
The Four-Question Strategy provides a framework for developing student ideas in coming up with a topic and a problem question for an experiment. A completed example of the Four-Question strategy is provided below after a brief description of the purpose of each part of the strategy. A blank copy is provided for students.
Four Question Strategy Worksheet
| Question 1 | Asks the student to think of a general topic in which he or she is interested and what materials are available that could be used in an experiment. |
| Question 2 | Asks the student to list what he or she knows about how this subject acts. |
| Question 3 | Prompts the student to think about how the materials listed for Question 1 could be manipulated/changed to affect the action (Question 2). The item the student chooses to change will become the independent, or manipulated, variable. The other potential variables on this list will become the constants. |
| Question 4 | Asks the student to think about how a response to the changed variable can be measured or described. This process defines the dependent, or responding, variable for the experiment. |
Sample Completed Experimental Design Diagram
Blank Experimental Design Diagram
One way to assure the title accurately describes the experiment is to use this form:
“The Effect of
(Independent Variable) on (Dependent Variable).” The teacher decides whether or not this form is insisted
upon. However, students preparing for science fair competitions are encouraged to choose a title that
will grab the attention of the judges; in such cases it is suggested that a subtitle be in the form “The
Effect of….”
The hypothesis predicts the relationship between the independent variable and the dependent variable and must be written as an “If …, then …” statement. The student should write the independent and dependent variable on the diagram before writing the hypothesis. Then the process of completing the “If…, then….” statement can be implemented in this way:
“If the (IV) is (how the IV is changed in the experiment), then the (DV) will (describe the predicted effect).”
Students often avoid describing the prediction of the effect by using words like “change” or “vary” rather than thinking through how changing the independent variable will cause the effect. If they do leave out a prediction, then they don’t have a meaningful hypothesis with which to compare their actual results. It is suggested that after class members have written hypotheses for an experiment the hypotheses be read aloud so that each one may be analyzed during a class discussion. In this way students will practice identifying good examples and will learn to recognize hypotheses that do not identify the variables or predict results.
| Levels of the IV (Label here the level that will serve as the control, if there is one.) |
|
| Repeated Trials |
The independent variable, or manipulated variable, is the one of several possible variables the investigator has chosen to change. Referring to the Four-Question Strategy, it is the one variable chosen from the answers to Question 3 (How can I change the set of _______ materials to affect the action?) that the student has decided to test. All the other possible changes listed must be kept as constants for this experiment. When the student decides how the independent variable will be changed, these changes are listed as levels of the independent variable. Vertical lines should be drawn within the “Levels” box to make a smaller box for each one. These vertical lines should be continued down through the box for repeated trials.
When the student designates the levels of the IV, it is also time to label the level which will act as the control, if there can be a control. There are three ways to consider how to designate a control. (1) In many experimental situations the factor that is the independent variable can be omitted and that level of the variable will be the standard of comparison for the other levels, called the control. If the purpose is to determine how fertilizer affects plant height, one level of the variable should be to omit fertilizer, “no fertilizer,” and all other levels will be compared to this control. (2) For some experiments, there is an outside standard that may be used as the control (e.g., the recommended amount of fertilizer printed on the bag). (3) In many experiments students choose to do, comparisons are made and it is impossible to omit the independent variable, so there is no control. Some will advise that the experimenter may select the level that will be the control; others feel it is best to say that for experiments that are merely comparisons and omitting the variable is impossible, there will be no control. For instance, if the experiment is to compare the effect of different soil types on plant growth, soil cannot be omitted, so there is no control.
The number of repeated trials must be the same for each level of the variable, and for a classroom situation the number of repetitions is limited by time. If several lab groups are doing the same experiment, each group’s data may be considered as a repeated trial to be averaged with the data from the other groups. The purpose of repeated trials is to reduce the effect of errors, so the more the better. As a guideline for individual experiments, students should do a minimum of three trials for each level. Since the diagram is a planning tool that allows students to visualize their thinking when they are designing an experiment, the box under the independent variable should not be transformed into a data table. Students must draw a data table to correspond to the requirements for recording data for their experiment.
The dependent variable is the variable that responds to the independent variable. It is usually measured, and the unit of measure should be indicated. In some cases the dependent variable may be counted or observed objectively. It may be necessary for the teacher to work with students to develop specific descriptions or a scale with which to compare qualitative results. For example, if a color change is being measured, a color chart can be used for comparisons.
To be sure that the experiment will produce valid data related to the effect of changing one independent variable, all other potential variables become constants. All the possible factors that could have been changed have been identified in Question 3 of the Four-Question Strategy. Each of these factors must be specified exactly. It is important to emphasize repeatedly when teaching students to plan an experiment that what they write must be so clear that all people who read what is written will know how to do the experiment in the same way. For instance, it is not enough to say “water” is a constant: the student must say how much water, what type or source of water, and when the water will be applied.