Polling Terminologies Explained.

Polling is a scientific way of collecting and analysing opinion in order to find out how the
public feels about a subject, person, group or event. To understand polling, you need to
know some basic terms. Here are a few.

Beginning and Ending Dates
This refers to the period in which the survey was done. It’s important to know when polling is
done so you can place the responses in context, especially when the results can and may
have been affected by events and also to track changes in views in cases where the same
questions are repeated in multiple polls. Our survey on the 2019 budget statement for
example, was taken between 19 th and 25 th November, 2019, four days after the statement
had been delivered in parliament.

Dataset
Datasets consist of all the information gathered which need to be analysed. Typically, this
will include the answers to specific questions asked in the survey as well as information
about the respondents. It is this information which is then laid out in graphic form to tell us,
for example, which percentage of respondents think a national cathedral is a good idea and
in which region or age bracket, it is most popular.

Feeling Thermometer
This is a type of polling question where the respondent is asked to gauge the degree of their
like or dislike towards a person, event or topic. For example, we often ask respondents to
rate their feelings towards something on a scale of 1 – 5, with 5 typically being the most of
the highest degree.

Filter Question
A type of question used on surveys in order to determine which subsequent (if any)
questions to ask. This determines the relevance to the respondent. For example, a poll on a
party’s internal primary would seek to determine whether the respondent is eligible to vote,
determining the weight to place on their responses.

Interview Method/Mode
An indication of the method or mode of interviewing: mail, telephone, or in-person. A poll
may focus exclusively on one method or a mixture of two or more. iPollgh uses the online
method exclusively, with respondents taking the poll either through the mobile app or the
website.

Population
The theoretical population from which the sample was drawn. In Ghana, polls based on the
general adult population would thus have a population of over 29,000,000. Responses would
be measured against this number. Other polls for specific groups, such as students, women
or pensioners, would have a population size that reflects the number of people in the country
who fall into that group.

Reliability/Confidence level
The quality of measurement that suggests that the same results would be repeated each
time in repeated observations of the same phenomenon. In other words, it expresses the
level of confidence that where the poll be taken by a different group of respondents, a result
close to it would be realised. In our Eye on the Big Issues Poll, for example, we assert that
we are 95% confident that the results would be repeated with a comparable but different
sample.

Sample size
This is the total number of random eligible participants who participate in the survey. The
desired sample size is determined by the necessary statistical quality for the survey results.
For our first government approval ratings survey, 1035 people took the survey, making that
the sample size.

Margin of Error
The margin error is an estimate of how a sample statistic is expected to deviate from the
population parameter. For instance, a margin of error of 3 percent indicates that a survey of
the total population would produce a result that varies no more than 3 percent more or less
from the results achieved with the sample.

Sampling
A method of selecting elements (or units) from the target population in a way that is
representative. Types of sampling include: Simple random sampling, stratified sampling,
systematic sampling, and multi-stage cluster sampling.

Weighting
Weighting is a technique used to reflect differences in the number of population units that
each case in a dataset represents. This means adjusting your units so that they offer a true
demographic reflection of the population. Polling organizations may have different methods
for their weighting procedures, but it generally involves the multiplication of survey
observations by one or more factors in order to increase or decrease the emphasis that will
be given to the observations when analyzing the data.