| Sumario: | Predictability is also partly a testable hypothesis and findings of many studies clearly attest to the predictability of human interaction. Sociology clearly has a subject matter which is amenable to scientific study. The purpose of the paper is to analyze a subtle but fundamental noninstrument source of indeterminacy in attitude measurement and to consider some of its implications. The data for all sciences consists of human responses to environmental events. The general procedure for minimizing indeterminacy in measurement due to the limited discriminal capacities of the human as a researcher is to have independent observers check on each other's findings. The precision of measurement is being operationally defined in an unfamiliar sense. It ordinarily denotes something like sampling error, the items in the scale are regarded as a sample of all possible items in a given attitude domain. The accuracy of best estimate is increased by increasing the number of items in the scale, thus increasing the number of endorsements.
|
|---|