Posicione Pesquisa de Mercado - Qualitativas, quantitativas, focus group

When you are collecting data on an election or a response to any other form of research marketing, research can be an effective tool to measure the response of voters / consumers issues. As we know, specifically in the electoral context, different companies / institutions investigate and collect data on the same phenomena. And why different numbers? One answer lies in determining the margin of error.

We now present a step-by-step that will help you understand how margin of error is recorded in research reports.

Instructions:

Step 1^o: Identify the margin of error presented in the research. Suppose that three per cent, more or less. The existence of a margin of error is inherent in the sampling process. Is statistics. Arises from the theory of probability. In selecting a sample for research, several factors prevent you from having a margin of error of zero. The principal is the heterogeneity of the population, ie, their differences are they demographic, psychographic and / or geographic. Remember, different people think differently.

Step 2^o: now how can we correctly interpret a survey of three percent error? The correct process to add and subtract the margin of error value obtained in the survey. Thus, we believe that the outcome of a given phenomenon can not be represented by a point, but by a range, which has a size of exactly twice the error in the survey.

Step 3^o: Exemplifying. Imagine that the candidate Y has twenty percent of intended votes, the candidate X has seventeen percent and the candidate Z twelve percent. A simple reading of the numbers give the following information: candidate Y first, second X and Z in third. Analyzing the situation in a statistical way we read the following. X and candidate Y tied within the margin and candidates Y and Z also tied within the margin. The correct reading, because the latitudes or ranges of the candidates are as follows.

Candidate Y: 20% (range, your result is between 23% and 17%)

Candidate X: 17% (range, your result is between 20% and 14%)

Candidate Z: 12% (range, your result is between 15% and 9%)

Thus, it is how we should proceed to read quantitative research to adopt the random sampling process and indicate their respective sampling error.

And one last observation: the sampling error is the error that the researcher incurred when selecting a sample and not the entire population. There is no way the error of weak implementation of the research process (more specifically the process of data collection).