Scientific research generally measures something, and often compares that measurement to something else. It is usually not possible or practical to perform the measurement on whole populations (for example, all women between the ages of 25 and 40). Therefore, samples are used to represent the whole population, and the measurements are made on those samples. The measurements are then analyzed, and conclusions are drawn. If the sampling was done appropriately, we can assume that what is true for the sample is true for the whole population.
The Research Question
Research questions are generated from different sources. An observation might lead a researcher to formulate a question, a review of what has been published on a given topic may point out an area that has not been investigated, or the results in one study may lead to further questions to be answered. For example, many health food supplements contain the ingredient XYZ. Some people think that XYZ has been causing weight gain in people regularly taking supplements containing XYZ. Dr. Smith wants to determine if rats fed XYZ gain more weight than rats eating their normal diet. Her research question is: “Will a diet containing XYZ make rats gain weight?”
The hypothesis is a prediction of the results of a study; it is a statement of the answer the researcher expects to find to the research question. Dr. Smith’s hypothesis is: “During the study, the rats that receive XYZ in their diet will gain more weight than rats not receiving XYZ.”
In order to determine if her hypothesis is correct, Dr. Smith plans a study that will allow her to determine if the rats receiving the supplement gain weight. Additionally, she must collect data on the weight of rats that do not receive the supplement so she can compare the two groups. It is important that the two groups be as similar as possible at the beginning of the experiment, so that any difference in the weight at the end of the study will be due to the supplement and not to some other difference that existed between the two groups. Dr. Smith must also ensure that the groups are treated exactly the same way during the study, so that the only difference between them is the presence or absence of the supplement in their diet. The experimental design describes in detail what treatment will be used, how much, how often, and so forth. It also describes what, when, and how measurements will be taken.
Common Scientific Research Terms & Concepts
A variable is anything that can potentially affect the results of the experiment. For example, in Dr. Smith’s study, temperature is a variable that can affect the results. If some of the rats are kept in a room where the average temperature is 64 °F and others are kept in a room that averages 79 °F, the difference in temperature may alter the growth rates in ways that are unrelated to the dietary supplement under study. Other variables include the gender, age, strain, and health status of the animals.
The independent variable is the variable that is manipulated by the investigator. In Dr. Smith’s experiment, the independent variable is the dietary supplement that will be added to the diet of the rats in the experimental group.
The dependent variable is the variable that is measured during the study. Dr. Smith will weigh her rats at the end of the study. Weight is therefore the dependent variable in this study.
In simple terms, the independent variable is what you think will cause the change that you described in your hypothesis and the dependent variable is what you measure in order to determine if a change has taken place.
Variables are tested by using an experimental group and a control group. The experimental group is the group that receives the treatment (the dietary supplement XYZ in Dr. Smith’s study) in a scientific study. The control group is a group of animals that is as similar as possible to the experimental group, but that does not receive the treatment. In Dr. Smith’s study, the control group received a routine diet without the XYZ supplement.
How many animals are to be included in the study (the sample size) will depend on the type of study and the level of confidence that the investigator determines is needed. If you could include a very large number of rats in your experiment, you would be very confident that the results could be applied to the whole population of rats. Because of the 3Rs principle and other practical considerations, the investigator must determine the minimum number of animals needed to obtain results that one can have confidence in. For Dr. Smith’s study, an important question will be: “Is this a true representation of the effect of dietary supplement XYZ on rats?” You cannot answer “yes” to that question with 100% confidence unless a very large number of rats have been tested, which is not a practical alternative. In reality, the sample size is as small as possible while providing enough statistical power to obtain conclusive results. There are methods that can be used to determine what sample size should be used to provide statistical confidence in the interpretation of the results.
To ensure that any effect demonstrated during the study is due to the independent variable and not other influences, it is important to make sure that the experimental group and the control group are as similar as possible at the beginning of the study. For example, in Dr. Smith’s experiment, all rats should be of similar weight at the beginning of the experiment. Other factors to be addressed are gender (males usually grow faster and larger than females), age (older rats don’t gain weight at the same rate as younger rats), and stock or strain (Sprague-Dawley rats generally grow larger and faster than Long-Evans rats, for example). If Dr. Smith intends to prove that the supplement is the reason that the rats in the experimental group gain more weight, then she must start with rats of the same weight, age, gender, and stock or strain. If these factors are not controlled, they could influence the results of the study.
What would you study if you were going to do research?