Opinion polls are a specific type of survey conducted using human subjects and are designed to gather a population’s opinion or views on a particular subject or controversial issue. Commonly, opinion polls are utilized for predicting election results, displaying the public’s opinion on economic or social issues, surveying religious beliefs and so on.
When conducting an opinion poll, many factors need to be considered. Firstly, the sample size and sample composition can greatly affect the results of an opinion poll. For example, if the sample size is too small, the results gathered will be far from representative of the total population being considered. Furthermore, sample composition is also critical to ensure that an accurate cross-section of the type of persons in a population is being analyzed. The results will be biased if only one type of individual were sampled for a poll considering the whole population. Lastly, constructing clear and strong questions, which cannot be misunderstood, is also imperative. Sometimes, simple yes or no questions, with an undecided option, are the best question format.
The use of the controlled sample method and the area sampling method are techniques commonly utilized for controlling sample composition. A controlled sample attempts to represent proportionately all of the different types of a people in a population. For example if a population was comprised of 30% farmers, 40% government sector workers, 20% private sector workers and 10% unemployed persons, these percentages would need to be used to decide the proportions of the total sample which would be from different occupational sectors.
An area sampling method differs because it does not compile a cross section based on persons, but rather on geographical location. For example, sectioning would consider whether individuals were from the north or the south, or from big cities versus small towns, for example.
The concept of an opinion poll is rather simple to comprehend. However, there are several factors which need to be considered to allow for accurate and effective surveys to be compiled.