Davis Chapter 14: Experimentation Essay Examples

There is a time when advertisers need more than descriptive research to understand consumers current behaviors and attitudes. They sometimes require an understanding of cause and effect, which is obtained through experimental research in which an advertiser alters something in the consumer’s environment, or product in order to see what would happen. The goal of an experiment is usually to determine the effect of changes in one area on one or more other areas. An experiment is used to vary the amount of advertising placed in particular markets to determine how advertising weight affects target audience, advertising awareness and message recall. An experiment answers the question of “to what extent the amount of advertising exposure affect advertising exposure affects advertising awareness and message recall?”
Surveys and other personal interviews are good methods of collecting information on consumer’s attitudes and, beliefs but in certain instances advertisers need to understand those individual behaviors. Surveys are used in instances where the respondents are asked to recall and report on their behaviors while experiments are used to bypass recall and observe behaviors. On given times, survey results can be less reliable for explaining behaviors hence the need for experimentation. Unreliability can be so when questions require recall especially over long periods of time, when individuals are asked for introspection in areas of behavior that they less think about and when individuals are asked to use memory in the attribution of cause and effect. Surveys can also be less reliable in cases where there may be complex, multiple and interrelated influences on behavior. In such cases, experimentation is usually the preferred option.
Experiments are most appropriate when one needs to understand the effect of changes in one area on other areas (causation). Experiments consist of four basic steps that include: identification of what need to learned, manipulation of variables, observation of the effects of manipulation on other variables and the determination of the extent to which the observed effects can be attributed to the actions taken. In the case of Peter Pizza’s experiment, Peter decided on understanding the effect of water on crust texture, varied the amounts of water on the recipes, observed the effect of water on the crust’s texture and decided on water on crust C was the best. Every experiment has an independent variable, dependent variable and manipulation. A variable is a thing that needs measuring by the experiment. The independent variable is what the experiment manipulates. In the case of Peter’s Pizza, the independent variable was the amount of water. The dependent variable is what the researcher aims to explain and as a result used as a measure to evaluate the influence of the independent variable. Dependent variables in Peter’s experiment were the crust’s firmness, color and crispness. Independent variables should manipulate in a systematic way. Peter had variations for water. Independent variables cause changes in dependent variables and vice versa.
Experiments try to explain casualty, which is that changes in independent variables cause changes in dependent variables. The hardest task is usually to prove casualty in a convincing way without having problems. A well-designed experiment would permit one to attribute changes in the independent variables caused changes in the dependent variable. The criteria need to be followed in order to establish such a relationship between the variables. The criteria include making events take place in a certain order, events taking place concurrently and showing an explicit relationship, reducing and eliminating any alternative explanations and having the strength of associations. Having the events running at the same time and showing an explicit relationship is the most crucial criterion as it affects the experiment’s internal validity. An experiment’s internal validity refers to the extent at which one can eliminate alternative explanations for the experimental results observed. The higher a researcher’s ability can show the correlation between manipulations of the independent variable to the dependent variable, the higher the level of internal validity of the experiment. It also boosts the confidence a researcher has in inferring causation.
The internal validity of an experiment can reduced if there can be alternative explanations. Researchers have identified a number of threats to experiments internal validity. These threats include pre-measurement, interaction and testing, problems due to data collection (instrumentation), problems associated with the chosen sample (maturation, selection and mortality), problems caused by the study context (history) and problems attributed to the researcher (bias). These problems negatively affect the experiment in the sense that they raise alternative explanations on the relationship between the variables and hence decreasing a researcher’s confidence in drawing conclusions from the research.