It is useful to design an experiment in certain reception and audience studies, to see how the audience responds to certain stimuli, and to validate certain types of hypotheses, especially relational ones. Experiments aim to reproduce in the laboratory some of the conditions of the natural or, in our case, social world, with the advantage that by proceeding in this way it is possible to isolate and better control the variables and their mutual relationship, while in the natural or social world there is another series of conditioning factors that cannot always be observed or considered.

The control of variables is another of the great advantages of experiments. On the other hand, when the aim is precisely to take the context into account, another type of method is recommended, for example qualitative observation.

Although experiments are normally designed and carried out in a laboratory, where the variables can be isolated and controlled, there are also experiments that are carried out in the field, and aim to record, or at least observe, the possible influence of other circumstances that may influence the result. The difference between the two, therefore, is the control of conditions.

Other advantages of designing an experiment are the identification of the causes of certain phenomena, in our case social and communicative, the evaluation of the effects of certain variables, and the precision with which the mechanisms in which the variables operate with each other can be described. John Brewer and Albert Hunter put it this way:

Experimentation, more than other styles of research, promises clear causal inferences. Its strategy is to manipulate exposure to an hypoteshized cause, while controlling for the contaminating influence of other possible causes by the use of control groups and by the random assignment of subjects to control an experimental situation (Brewer and Hunter, 1989).

There are basically three types of experiments in terms of design:

1) A person or group, a single variable.

2) Two groups, a pair of variables (dependent / independent).

3) Factorial: More than one group and more than one pair of variables. They are, it goes without saying, the most complex and time consuming but the ones that yield the most elaborate results.

The experiments can be projected or ex-post-facto. While the first seeks to determine the cause-effect relationship between variables, the second type of experiment seeks to reconstruct a certain social or communicative situation, for example, to reconstruct the reactions to viewing certain types of scenes in a television series.

The experiments can be carried out simultaneously for all groups or longitudinally, over time so that we can also record the variations that may occur through the application of new conditions, for example the experience acquired from the user and the development

When designing an experiment there are three basic elements to take into account:

1) A group of people who undergo a stimulus to measure their responses.

2) A stimulus, usually the independent variable or cause of the dependent variable.

3) A transformation that can be registered and, where appropriate, measured, which is the independent variable or, as it were, the effect.

Normally, when an experiment is designed there are at least two types of groups that are subjected to it:

The control group and

The experimental group itself.

The control group is the one whose reactions are known, while the experimental group is the one in which we intend to measure the reactions.

Both groups are equal, the difference is the stimulus to which they are subjected.

The techniques that can be used to present and measure the stimulus are diverse. These include providing participants with a list of adjectives to choose from to rate the response to the stimulus, or to grade it.

Another technique is word association and concept classification. On other occasions, the participants are asked personal constructions or even narrations (storytelling), to narrate their experiences once the stimulus has been presented to them. Another method is to invite participants to express their reactions out loud, or to play a role, or to fill out questionnaires. Finally, another technique is the simulation of certain social situations.

There are four methods to determine the cause:

1) The matching method.

2) The difference method.

3) The residual method.

4) The concomitance method (increase or decrease).

The experiments can be subjected to spontaneous conditions, and are called free experiments, or be subject to special conditions, and stimuli specially designed for the experiment. For example, in an experiment with Eye Tracker, a computer program that detects and records eye movement on a computer screen, a group of people with certain characteristics can be asked to read a real online communication medium or design a news item that gathers those variables that we want to consider.

The validity of the experiments is achieved by basically applying three techniques:

1) The parallel test (alternative-form method): The same result is sought using different methods on two occasions.

2) Internal consistency (split-half method): dividing the group into two halves to see if both offer similar results, and applying a correlation coefficient.

3) The stability test (test-retest method): The same experiment is repeated at various times in the same group to verify that it produces the same results.

One of the main problems of the experiments is representativeness. In fact, this is a problem common to many techniques: how a sample can reliably represent a certain universe, sometimes much larger, and how to combat randomness. A possible solution is to define well the profiles of participants from strata of the universe, in such a way that invariability and homogeneity are considered, which leads to looking for pairs of participants: men / women, young people / elderly people, etc. In addition, it is necessary to take into account both the possible disturbances and the possible appearance of new variables, if they are needed.

Other problems have to do with the biography itself and the personal singularities of the participants, which are difficult for them to respond absolutely to the defined type; the maturation and experience that the participants accumulate, which is detrimental to naturalness; the different selection and composition of the groups; and abandonment from the first session due to lack of motivation or interest, so it is advisable not to carry out many sessions with the same group.

Like any other technique, the experiments provide benefits but show some weaknesses, criteria that must also be taken into account when deciding on this technique:

Source: Own elaboration