Many are times one needs to collect data. There are several reasons why a survey might be conducted:
- To produce analyzed information towards a certain field
For instance, we can perform a survey on the number of diabetic patients suffering from hypertension. This data can be analyzed and stored for reference purposes.
- To provide customer feedback on a brand
We can also perform a survey on customer satisfaction on a specific brand and use this data do plan future marketing strategies.
- For training purposed
We can also perform survey to demonstrate how they are conducted and how data is collected and stored.
- As part of research
There are times a survey is conducted while trying to learn about a specific field. This type of survey mostly does not have a guarantee of what type of data or its distribution to expect.
To cost, time use and ensure validity of data, data is collected from a portion of the entire population and assumed to be a representation of the entire population.
When planning on conducting a survey we need to first determine:
- Method of data collection
- Size of samples
- Tools to use
Survey Methods
There are two main methods of performing surveys. These include:
- Systematic method and
- Random method
Systematic Method
At this point it is clear that the survey is conducted on a fraction of the population. This method gets a small group of people from the population and gets data from every member. Therefore, how many people should take your survey for reliable results determines the size of the groups.
Suppose we need to know the number of diabetic people in a county. We can determine to group people evenly from every point in the county. The people can be tested for diabetes and the fraction that is positive used with the total population
Random Method
This method collects data from a few members of the population at random. Still using the same example, we can test people at random for diabetes and use the positive fraction with the total population.
Survey Tools
The first decision is selection of the tools to use. There are two tools that are very common in conduction of surveys:
- Interviews
- Questionnaire
Interviews
These are verbal questions on interviewees to get comprehensive data on a specific topic. The questions can have follow up questions to express the data further.
Questionnaires
This involves a written set of simple – mostly yes/no- questions.
Theyare prone to several shortcomings:
They can be quite time consuming.
The population being targeted might not all be literate.
The paperwork can even get lost or destroyed in the process.
It is also very difficult to ask reasonable follow up questions.
Size of samples
If you are conducting a survey of people, this can determine how many people should take your survey for reliable feedback.
Consider a medical survey on how many people are diabetic in a county. There are several factors that determine the size of the sample to test. These include:
- The size of the population
- The budget at hand
- Time available
Size of the population
The total size of the population is a very important factor. Firstly, we cannot have a sample that exceeds the total population. Secondly, we will come to note that as we increase the size of the sample the accuracy increases.
From example assuming the county has a total population of 100,000 people. We can take a sample of ten people and maybe all turn out negative. We can conclude that nobody is diabetic.
But if we use a sample of, say 100 people, and find 7 people as positive, we can conclude that 7,000 people are diabetic.
Increasing the sample to maybe 1,000 people and finding 73 people positive, we can conclude the that 7,300 people are diabetic.
It is therefore advisable to maximize the size of the samples
The budget
It is also very important to consider the budget. We have to consider that there are several resources and hours of man power that need to be funded. This is one of the factors that limit us from using the entire population as the sample.
Time factor
The amount of time available to perform the survey is also quite important since the data needs to be collected to completion within the set timeframe. This also limits us from using the entire population as the sample. For instance, we might have come across times when the census does not manage to cover the entire population due to the timeframe.
Therefore, there are a number of factors that need to be put into consideration when determining how many people should take your survey for reliable results
