Businesses, educators, government officials, and ordinary people have a need or interest in gathering information. A survey is just that: a way to gather information and learn from respondents. Although surveys might seem easy to design at first glance, there is more to them than meets the eye. Read on to learn how to produce the best and most useful surveys and make your life easier.
Steps
Part 1 of 3: Design the survey
Step 1. Set your survey goals
In short, what do you hope to get out of the survey? The questions you ask should be directed at this essential idea.
- For example, let's say you are an employer and you want to decide if your employees are happy. The questions you ask in the survey should focus directly or indirectly on worker happiness. You could ask them directly "On a scale of 1 to 10, how happy are you at your job?" Or you could ask a more indirect question, such as "True or false: I wake up every morning feeling that my work has a purpose."
- Once you have designed all the questions in your survey, it is advisable that you review each of them and ask yourself how they will help you achieve your goal. Eliminate any questions that do not provide essential information about the purpose of the survey.
Step 2. Take some time to think about how to get the most honest answers
If your goal is to find out if your employees are happy, you want an honest answer. In fact, you will always want honest answers to surveys, but you may find it difficult to get them if your employees feel that they could lose something, such as respect, position, etc., if they answer honestly. Consider whether or not you will need to fine-tune the survey in order to get accurate results. For example, in the case of an employee happiness survey, you may want to give them the option to respond anonymously.
Step 3. Decide on the best survey method to obtain valuable information
Some of the options include telephone surveys, personal interviews, mail-in surveys, or online questionnaires. Each survey method has advantages and disadvantages that you should consider based on budget, available personnel, and other resources.
- In general, personal interviews, while time-consuming and expensive, provide the most representative results and the most detailed responses. On the other hand, online questionnaires sometimes have the most significant margins of error, but they are the simplest and cheapest surveys that can be done.
- If you are only going to support yourself with one form of survey, such as an online questionnaire, consider surveying more people to compensate for the likely margins of error. To get the most accurate results, you may want to run several different types of surveys.
Step 4. Think about ensuring the accuracy of the survey
A survey that involves one or two interviewees will tell you something about each of them, but it will give you almost no accurate information on trends. In order to decide how many people you should survey, you will need to have two pieces of information:
- The size of the target population. What segment of the population do you want to understand? If you want to know the percentage of happiness in your company, your population will be the size of your company. If you want to study the adoption of condom use in Uganda, your population will be the size of the number of inhabitants that Uganda has, approximately 35 million.
- The assurance that your results will be accurate. People talk about two things when it comes to the accuracy of a survey: margin of error and confidence interval. The margin of error is the degree of insecurity you have in your survey results, the confidence interval is the degree of confidence you have that the survey samples accurately represent the population.
Step 5. Select your sample size based on your target population and desired level of precision
Once you have responded to the previous answers ("What is my target population?" And "How accurate do I need the results to be?"), You can start thinking about how many people you need to interview in order to get the desired results.. In the table below, select your target population on the left and then your margin of error to estimate how many surveys you need to do. As a general rule of thumb, the more surveys you take, the lower your margin of error.
| Population | Error range | Confidence interval | ||||
|---|---|---|---|---|---|---|
| 10% | 5% | 1% | 90% | 95% | 99% | |
| 100 | 50 | 80 | 99 | 74 | 80 | 88 |
| 500 | 81 | 218 | 476 | 176 | 218 | 286 |
| 1, 000 | 88 | 278 | 906 | 215 | 278 | 400 |
| 10, 000 | 96 | 370 | 4, 900 | 264 | 370 | 623 |
| 100, 000 | 96 | 383 | 8, 763 | 270 | 383 | 660 |
| 1, 000, 000+ | 97 | 384 | 9, 513 | 271 | 384 | 664 |
Part 2 of 3: Asking the Right Questions
Step 1. Decide if you will use suggested answer questions, spontaneous, or a combination of both
How much do you know the respondents? What is the purpose of your survey, to collect data on ideas you already know, or to explore new ideas? If you are collecting data on ideas that you already know, you will probably want to lean on suggested answer questions. If you are exploring entirely new ideas, you will probably want to lean on spontaneous response questions.
-
Suggested answer questions:
asks a question and below offers answers to choose from. An example of a suggested answer question might be:
(1) "What is your favorite online activity?"
(a) Chat, instant messaging (b) Social networks
(c) Share knowledge, forums
(d) Buying, e-commerce
-
Spontaneous response questions:
removes the default responses from the survey. Rather than guiding the interviewee in a particular direction by offering answers to choose from, spontaneous response questions encourage the respondent to develop a very personal response. The following is an example of a spontaneous response question:
(2) "Talk about your experience when you first walked into an Apple Store."
Answer:
Step 2. Choose partially suggested answer questions to gather a little more information and obtain data that can be analyzed
The downside of suggested answer questions is that they are often not very specific. The downside of a spontaneous response question is that it is difficult to analyze the answer and put it on a spreadsheet. The partially suggested answer question softens the drawbacks of both:
(3) "How would you describe your attitude towards paying for music? Check all that apply." (__) I never pay for music. (__) As a general rule, I pay for music I listen to. (__) I often download music illegally. (__) I rarely download music illegally. (__) I could be persuaded to pay for music if they gave me something else in return. (__) Nothing could convince me to pay for music. (__) I feel bad for musicians who try to earn a living with their compositions. (__) I don't feel bad for musicians who try to earn a living with their compositions.
Step 3. Ask a "matrix" or "ranking" question
This is a subgroup that follows from the suggested answer question. The goal is to probe how respondents would rate their experience on a scale. Your scale can be numerical or a more sophisticated rubric:
(4) "The Brooklyn Zoo is fun for kids and adults alike." (a) Strongly disagree (b) Disagree (c) Agree (d) Strongly agree
Step 4. Ask a ranking question to get an ordered list of preferences
With a rating question you will achieve a better result than with a rating question when it comes to obtaining the opinions of the respondents on a given topic. The following is an example of a ranking question: (5) "In the spaces below, rate the brands you trust the most, with '1' being the most trustworthy and '5' being the least trustworthy." (a) ___ McDonald's (b) ___ Google (c) ___ Walmart (d) ___ Costco (e) ___ Apple
Step 5. When developing suggested answer questions, add an option that includes all the others to the end of the answer set
It might be helpful to include options like "Other", "None of the above", etc., at the end of each group. These options generally help you get more accurate answers. Without an inclusive phrase, the respondent who does not find the answer that fits what he thinks will be forced to choose an imprecise one in order to complete the questionnaire.
Part 3 of 3: Distribute the survey
Step 1. Find a way to distribute your survey
Once you've decided what type (s) of survey (s) you will use, you will want to consider how to distribute it to your respondents.
- The Internet has facilitated the preparation, design and delivery of questionnaires. Services such as Google Forms, SurveyMonhey, and others, offer easy-to-do surveys and their services are free.
- If you are distributing telephone surveys or want to conduct personal surveys, consider that you will have to invest money. The data you collect will usually be more representative, but this comes at a price. You can usually hire professionals to conduct the interview for you.
Step 2. Make it as easy as possible to return the information
Including prepaid postage on mail-in surveys will increase your chances of getting them sent back to you. Distributing the survey at inconvenient times will discourage respondent participation. A group that must stay after work hours or at the end of a busy day could give you information distorted by fatigue or resentment.
Step 3. Analyze the results of the survey
If your data isn't all together in one location, this might be the time to put it together. Excel is a great tool to do so, use it to run formulas, create charts, and analyze data. In short, to determine what respondents said.
Step 4. Develop what you have learned and implement it
Now ask yourself why. For example, why are your workers unhappy? The answer could be embedded in the answers. If not, you may be able to create a new survey to help you answer those questions. Then, when you've figured out why ("My employees aren't happy because they don't get enough encouragement"), you can implement and reinforce your new strategy.