Empirical Construct

Accuracy – The degree to which calculations or estimations are in proximity to exact or real values. See also: Precision

Base Level – Prior to drawing conclusion from research data, it is frequently necessary to be informed about what is called the base level, conditions with which the research begins. For example, when, after an advertising campaign, 25% of a sample demonstrates positive buying intentions, that response does not necessarily signify that the campaign has been effective. If the conditions prior to the advertising campaign are unknown, it is not possible to make a statement concerning the effectiveness of the campaign. Was the buying intention prior to the campaign 0%, 10% or 24% These data are important.

Borderline – A case, occurrence or event that is difficult to judge or place. For example: should a person be classified in the highest or in one but the highest social class?

Calibration – Testing measuring and weighing equipment against specified standards and demands. When these specified conditions are met, a mark or certificate of approval is furnished (guarantee certificate). For example: to compare the results of a local study to the standards for the entire country.

Casuistry – The theory of the single case or event. The study of a number of separate cases and their conclusions. These conclusions are indications, directions. Generalizations are not permitted. See also: Case study

Criterion – Norm, measure or standard to compare to or test against. For example: in a particular sample 20% is younger than 34 years. The criterion is 35 %, according to the official census. See also: Objective criterion/ Census

Empirical Construct – Syn: Construct The naming of the phenomenon found in the study. The research results give reason to assume the presence of the phenomenon; therefore, the empirical construct is based on the findings of one, or more, research project(s).

Judge – Syn: Rater A person who judges others for a specific reason. Purpose: construction of a Thurstone scale or the determination of performance or “value” of a specific product or brand. Judges may be either laymen or experts. See also: Thurstone Scale/Interjudge Reliability

Measuring Instrument – Any device used for measuring. Every questionnaire, psychological test, gauge of physiological response, etc. is a measuring instrument.

Measuring Sensitivity – Syn: Sensitivity of Measuring Instrument Degree to which the measuring instrument is able to differentiate between the different characteristics of certain individuals.

Method – Literally: a procedure or process for attaining an object. Should not be confused with “technique,” which indicates a systematic procedure for solving a particular problem. See also: Technique

Modeling – A scaling-down of a complex situation whereby an existing system is simplified to a framework of relations between a restricted number of elements. Modeling occurs in all branches of science; it is an attempt to create order in the complexity of things around us. See also: Model

Objectivity – Manner of working and thinking in which one pays attention only to the matter under consideration (the object) and attempts to act or function without prejudice. Objective in a research study is an ideal that can, however, never be attained. (Man is subjective in his perceptions.

Simulation – A method that imitates reality in the form of a model. The aim is to arrive at statements concerning reality on the basis of the results of experiments with this model. This method frequently involves the use of computer. See also: Simulation of thinking/Model

Subjectivity – Related to a person’s feelings, ideas, plans, etc. Non- objective. Social scientific research and marketing research (insofar as it relates to people not to products) attempt to work in as objective a manner as is possible; however, there are two fundamental problems that may endanger the researcher’s objectivity:

  1. The researcher himself id human. Prejudice in relation to a specific research project cannot be permitted ( the researcher should not influence the study in such a manner that the results obtained are in accordance with the researcher’s personal desires). The researcher must remain neutral and objective; As such,a belligerent respondent should not be evaluated differently from a pleasant respondent.
  2. The respondent does not relate in an objective manner to a research project (dishonest answers, “colored” answers). Only a machine (computer) can be entirely objective. See also: Objectivity

Survey Weariness – The situation in which a person, during a specific (short) period of time, has been selected and questioned several times for marketing and opinion research, with the results that he or she refuses further interviews, exhibits poor motivation (fatigue) or is capable of “expert response.” Research has shown, however, that despite frequency of the applied random walk method of sampling, survey weariness is actually quite minimal. Some persons are exposed to a greater degree of questioning. Such people are usually those who are easily accessible. Others (for example, farmers living in isolated locations) are relatively inaccessible to such exposure. See also: Random-Walk Method

Technique – Specific treatment according to certain rules that leads to a problem solution, e.g., in multivariate analysis. The term “technique” should not be confused with “method,” which indicates a general treatment. See also: Method

Test Anxiety – The fear experienced by a person being tested or studied. Though there is no real basis for this fear, extreme test anxiety can influence the results of a research project in a negative manner. A slight degree of test anxiety occurs quite frequently and appears to exert a positive influence (person tries best, is adequately motivated).

Test Wise – Having been tested several times and, as a consequence, having attained a certain degree of wisdom” in connection with the test. Persons who have been tested frequently have become familiar with the “tricks of the trade” and. For this reason, show better test performance (within certain limits) than do persons who experience a test for the first time. See also: Learning Effect

 

Leave a Comment

Web Analytics