For human rights NGOs today, effective advocacy is often complemented by survey data. Thankfully, modern technology offers individuals and NGOs a wide range of options for conducting survey research. YouTube, online tools and instructional websites make it look straightforward. However, conducting a legitimate survey and properly assessing the data can be challenging, especially if you don’t know where to start. This article will lay out key tips and tricks for setting up and starting a survey.
- Step 1 – Identifying your quantitative research question
- Step 2 – Identifying a sample pool
- Step 3 – Why is sample size important?
- Step 4 – Choosing the correct survey style
Step One: Identifying your quantitative research question
To begin with, you need to first identify what you want to discover through the survey. If the organization wishes to establish whether a certain population feels positively or negatively about a new law or a recent event, then the survey structure needs to be centered around that question. In research methods this is identified as a research question The survey will always include one research question and it is up to you whether or not to include a type of hypothesis predicting what will happen within the study. To understand how to formulate a research question that fits the survey being conducted you can refer to the Laerd dissertation site.
Step Two: Identifying a sample pool
After narrowing down the question(s) your organization wishes to answer, you need to identify your respondents, or the population to undertake the survey. Choosing the type of sampling is key here: you must sample within a group or population you are knowledgeable about. The more you know about their cultural habits, language, and social cues, the easier it will be to conduct a survey and get accurate answers. It is surprising how often NGOs go in to a survey project with the wrong sample pool for their needs. Make sure you have researched the context thoroughly before proceeding.
Step Three: Why is sample size important?
Once you have chosen your sample population, the next step is to calculate the sample size. A mistake often made is assuming that everyone in the sample chosen will show up and participate in the survey Calculating the Right Survey Sample Size. Picking the right survey size helps assess the survey’s confidence level/interval and margin of error.For a survey to be considered ‘successful’, the confidence level should ideally be 95% and the margin of error 5%.
For example, if you are targeting a sample pool of 1,000 people affected by human rights abuses, then you can be confident that around 300 people will participate in the survey. The remaining 700 people may or may not participate. Furthermore, it can be predicted that around 100 will not participate due to unforeseen circumstances.
The most important insight to remember when choosing a sample size is that no number is too large. The more people who participate (fitting the respondent criteria), the stronger the data will be.
However, if the sample size is too low, the confidence level is lower than 95%, and the margin of error higher than 5%, then the effectiveness of the survey will also be compromised. This does not allow you to draw confident conclusions about your findings.
Step Four: Choosing the correct survey style
Once you are satisfied with your research question(s) and sample population, the only step left is to choose a survey that fits your question and survey method to ensure the highest turnout of respondents. A more detailed description of survey types can be found here.
When conducting a survey, a general rule to remember is that survey questions should be designed so that they both fit with what your survey is trying to answer and make sense to the sample pool. A more detailed description of question types can be found in an upcoming article on types of questions.
It is also important to take into consideration cultural differences when designing survey. If you are a native English speaker, or do not speak the local language of your sample population, then you need to take extra care to ensure the meaning of your questionnaire is accurately translated.
Finally, it is important to choose the most effective method of conducting your survey with the sample pool. If the survey is trying to gauge public opinion on an issue and has few questions, then street surveying can be a viable option. If the survey requires more time and is longer, then in-person surveys would work better. There are also a wide range of online tools available to help you both design and carry out a survey, while also saving time. While the most common tools used is either SurveyMonkey or Google Survey, other types of online tools can be found here.
Floyd J.Fowler Jr, Survey Research Methods, 5th ed. (Center for Survey Research, University of Massachusetts:SAGE Publications, 2013), 3-4.
Last Updated: April 28, 2018
Author: Haley Joy Herbig