Chapter One

Introduction
Surveys are a way to compile data from a group of people, but they can be more than
that. Surveys are also direct insight into your customers and information about how they
feel about your company, products, and services. How did your customers perceive
your most recent product launch? How do they feel about your company’s social media
presence? Surveys are a direct line to find out and glean valuable information about
your customers and your company.

But not all surveys are the same, and different types of surveys can provide different
insights. There are many types of surveys your company can choose from, and it can
be hard to determine which one would be the best for your needs. So to help you figure
out the right type of surveys for your needs, we’ll walk you through the most common
survey types, the benefits of using surveys, and how to get started today.

Survey programming is the process of creating computer programs that enable

users to collect information from a variety of sources. This is a very useful tool for
gathering data for research and analysis purposes. Companies use survey
programming in a wide variety of sectors such as management, marketing, and
sales. They use it to collect information from people about their likes or dislikes of
certain products and services. In this case, the user can send the survey
questionnaires through various modes such as email, and phone.

Survey programming refers to translating the survey requirements into a questionnaire

such that meaningful insights can be derived out of it. For market research, data needs
to be collected from many sources in a number of ways. Survey software gathers data
easily and provides comprehensive reports and analytics which can be used to make
better business decisions.

Survey programming is the process of creating computer programs that enable

users to collect information from a variety of sources

Survey programming is the process of creating computer programs that enable

users to collect information from a variety of sources, like text files or databases.
Companies can use this program for many different types of surveys, such as
opinion polls, exit polls, customer satisfaction surveys, and more!

Businesses use surveys in all types of industries including marketing research,

political science research, and market analysis among others. This is a very useful
tool for gathering data for research and analysis purposes. It helps the researcher to
gather information from the respondents by asking them questions in a systematic
manner.

1
The survey programming language provides various options to create
questionnaires. Businesses can use these questionnaires for conducting surveys via
phone calls or online interactions with participants. The survey decides the mode of
delivery and it depends on their needs and preferences.

Information needed before you can start survey programming

Survey programming is a complex procedure and requires a lot of attention to detail.
It also needs thorough research before starting the survey.
As a result, it is important to know the following things before starting a survey:
1. The purpose of the survey.
2. The target audience.
3. Survey design.
4. Survey questions.
5. Sample size.
6. Sampling method.
7. Data analysis and reporting.

Some main advantages of survey programming

Here are some main advantages of survey programming:
 It makes it easy to create questionnaires and surveys.
 You can use it to access different data sets from a single source.

It offers a great way to gather data and insights from different sources. Survey
programming is a great tool that you can use to build surveys, questionnaires, and
quizzes. It offers a wide range of features and options for creating surveys.

Additionally, it is a way to create customised surveys that companies can use in a

variety of ways. It can help you create surveys that are more effective, save time
and money, and increase data reliability.

The advantage of survey programming is that it creates customised surveys for

clients who have specific needs or questions. Programmers design surveys with the
client in mind, so it's more likely to meet their needs and expectations. It also helps
you save time and money by eliminating the need for manual data entry. This saves
time and reduces costs associated with hiring workers for this task. It also helps
ensure data accuracy because it's built into the survey itself.

Another advantage is that it allows you to create a survey in any language. This
means that you do not need to translate your questionnaire into another language,
but simply create it in the desired language.

Survey programming also allows you to track responses in real-time. If there are
any questions or problems with the survey itself, you can address them immediately
without worrying about missing important information or having to wait until the end
of the survey period.

2
 Why is survey programming important?
The importance of survey programming is that it is an effective way of gathering
information about what customers think about your products or services and how
they can be improved upon. By asking questions in a structured manner, the results
will give you valuable insights into how to improve customer experience. For
example, you could ask customers how easy it was for them to find out where to buy
your product online. You could then compare this with the actual number of sales
made through the website during that period of time. In this way, it would be
possible to determine whether there are any problems with your online presence.

Benefits of survey programming

 Productivity: By applying survey logic tactics, you can create a number of surveys in
a matter of minutes. It reduces the time you need to create questions. Just like by
applying logic in software programming, you can achieve more with less code, in
survey programming you can collect more data with relatively less number of
questions. One can use the question library, copy questions from the same survey or
a different survey to reuse the question set.
 Efficiency: The purpose behind using a survey software is to design, distribute and
analyze data smartly. Such software comes with rich features that reduce the time to
do market research. It reduces the effort drastically as compared to traditional means
of data collection. With survey programming, you can collect more information about
your target audience with less number of questions.
 Conciseness: If you use a survey creation software efficiently, the survey
questionnaire will be very concise. This also improves the experience of respondents
as they have to answer a lesser number of questions. There are many methods to
make the questionnaire concise using survey building software. With survey logic and
programming, it becomes possible to present the same question based on the
respondent’s answer. Thus, the survey creator doesn’t have to create different
questions for different scenarios.
 Simplicity: Concise surveys are simple yet powerful. Because of the way they collect
data, generating insights from data analytics gets very simple. With survey
programming software, you can manage the market research campaign with
minimum human intervention. For instance, you can schedule the distribution of
invitation email at a specific time. Also, you can create a question list in your preferred
language and launch a survey based on the respondent’s language.

How can market researchers use survey programming?

Survey programming can be applied in a number of ways to increase the effectiveness
of the survey. Market researches should make the most of the below features available
in the survey creation tool.
1. Skip Logic: Skip logic moves respondents to a different question or page on the
basis of their selection of an option in the current question. This feature can take
respondents to a different question in the survey or skip some questions. You
can define custom rules to create a path for each respondent based on their
responses. For instance, a hotel can ask customers if they enjoyed their meals or
not. In this case, the first question would be if he/she had meals or not.

3
Based on the answer, the respondent can be presented a question on the restaurant
or the hotel.

4
 2. Compound/Delayed Branching: With simple branching or skip logic, you
cannot program a survey based on responses to multiple questions. Also, the
logic is executed immediately with simple branching. With compound branching,
you can set multiple criteria on a single question. With delayed branching, you
can use responses of previous questions to decide which question should be
presented.

3. Quota Control: Suppose you want only 1000 responses for a specific question, then
you can set the limit of the number of responses with quota control. Once you reach
the limit, no further responses will be accepted.

5
4. Dynamic text/comments: When you want respondents to give descriptive comments
on the selection of particular answer option, you can configure question to have
dynamic text/comments. The respondents will be displayed a text box only if they
select an option that is configured.

5. Extraction: Many times it is required to drill down further to get more insights.
Extraction programming can help to display question options out of options selected
in the current question. This enables you to present the selected options of a multi-
select question as answer options of the next question.

6
6. Show/Hide Questions: Sometimes it is required to hide a question based on an
option selected. The only condition is that there must be a page break between these
questions. For instance, respondents from different countries can be asked different
questions. In this case, the survey administrator will create one list of questions but
set conditions so that people are presented only relevant questions.

7. Show/Hide Options: You can program the survey to show or hide answer options in
either a matrix, single-choice, or select-many questions based on predefined criteria.
These criteria could be answers to previous questions or custom variables assigned
beforehand.

7
8. Scoring: Surveys can be used to calculate scores in real-time. You can conduct
online tests and compute scores. Further, it can be configured to display the total or
section-wise score to the test taker immediately after the test. Or they can be sent
scores in a separate email.

9. Javascript Logic: If in any case, the above logic mechanism doesn’t satisfy the
business requirements, custom javascript logic can also be applied to the questions.
You can insert logic through javascript code either before the page loads or after the
respondent hits the submit button.

8
9
 Chapter Two
A Survey Programmer
A survey programmer is an individual that creates user-friendly surveys, translating
written questionnaires that are then used to collect participant data to achieve research
goals. Their efforts are a crucial step to get market research up and running.

Are you interested in a career as a survey programmer for the market research
industry? This field involves designing and developing survey software and data
collection systems to better meet the needs of research clients. Survey programming is
a highly specialized field that requires an understanding of research methodology, data
analysis, computer programming, and client service.

First, the survey programmer takes the requested information for a market research
project into an intractable survey interface that will capture the data from participants
and translate this into actionable insights for clients. This includes programming the
actions of certain question types, the survey logic, and putting into place any
redirections through various parts of a survey experience.

For example, if a questionnaire asks, “Did you eat dinner last night?” with answer
selections “Yes or No” and you answer, “No” it doesn’t make sense to then ask, “What
did you eat for dinner last night?” The survey logic, if set up correctly, would
automatically skip that question. Survey programmers ensure that the survey questions
flow smoothly. The insights and analytics that are then derived from the survey project
enable leaders to make crucial business decisions.
Survey programmers typically work in the research departments of corporations and at
specialty market research firms. Because they are responsible for executing designed
surveys, analyzing data, and creating reports for their clients, they must be able to
communicate effectively with clients, critically evaluate written questionnaires, and be
able to work both independently and in a team setting.
Survey programmers exercise elements of customer service to support their clients
throughout the survey creation and testing process, and even into survey fielding.
This process includes attending meetings to determine specific client needs, developing
questions, designing the flow of the survey, programming the survey questions,
optimizing the survey for web and mobile applications, attending launch calls, testing,
making revisions, and processing the information collected from the study.

Survey programming is a technical field that in some ways might even compare it to
web development. Successful survey programmers have a variety of the following
technical and soft skills.
Individuals in this role must have a strong understanding of research methodology,
programming languages such as HTML, XHTML, CSS, Python, and JavaScript, and the
latest software technologies.
Having familiarity with the top software being used in survey programming today will
make you an ideal candidate. These programs include:
 Forsta Surveys

10
  SPSS Data Collection (formerly Dimensions)
 Qualtrics
 Alchemer
 QuestionPro

Skills survey programmer must possess

There is an increasing demand for data analysts and survey programmers as more
companies gain access to online research and begin to rely on that research to
make decisions. Skilled survey programmers will never be obsolete in our ever-
evolving industry that embraces technological advances and the need for accurate data
collection and analysis.

1. Logic and Technical Skills

Programming languages are loaded with “if/then”, “and/or/not”, and “true/false”

statements as participants progress through the survey. Programmers need to be able
to read and construct very complex series of logical statements to accomplish the most
basic part of their job role. Additionally, as clients want more complicated designs
including media uploads, video plays, eye tracking and other technical advances in their
surveys, survey programmers need to have the technical expertise to integrate their
clients’ requests seamlessly.

2. Attention to Detail

Programmers need to be attentive to every detail of a survey instrument, both when

interpreting the requirements and then, in writing the programming script. Missing the
smallest detail can have large negative consequences and lose valuable client data or
burn participants. Attention to detail is also important for maximizing participant
experience. Say for instance, you’re taking a survey and it has broken logic, you’ll likely
drop off- it is an inconvenience and bad experience. This results in an incomplete
survey response for the organization and the likelihood that participants won’t want to
participate in any more surveys. Additionally, in cases where participants have already
seen part of the survey, you must exclude those respondent IDs when you refield your
project, and that process can be expensive and hinder your feasibility. The ability to
catch details that may seem minute or easily passed over in testing prior to survey
launch can save organizations a lot of time and money.

3. Communication

Communication is a vital element in any role, and survey programming is no exception.

Straddling the technical and non-technical worlds, survey programmers need to know
how to adjust their communication based on their audience. Having excellent
communication and presentation skills is beneficial, as they often must present their
work to clients and other research stakeholders. Most clients won’t be able to follow
technical jargon, so they need to be able to convey their ideas in a way that is digestible
for a variety of audiences.

11
In addition, survey programmers must have strong problem-solving skills and the ability
to work with various technologies. They need to be able to think outside the box and
produce creative solutions to complex problems.
Survey Programming Capabilities
1. Randomized concept testing.
2. Embedding third-party interactive Flash tools in traditional surveys.
3. 3rd party DB transactions, using any kind of audio/video files.
4. Complex logic involving manual rotation, de-rotation (required for easier data
interpretation).
5. E-Mail Surveys in ASP, SQLServer [or MSAccess] or PHP, MySQL platforms along
with mailing facilities to respondents.
Data Processing Expertise
1. Highly qualified team for Data Processing.
2. Programmers trained in Quantum.
3. Highly qualified team for Data Processing.
4. Familiar with data formats and deliverable requirements.
5. E-Mail Surveys in ASP, SQLServer [or MSAccess] or PHP, MySQL platforms and
mailing facilities to the respondents
Experience with the following applications:
 Quantum
 SPSS
 WinCross

Services Offered for Data Processing, Analysis & Output Preparation

Output Preparation Data Analysis Data Processing

Insight Generation Cross Tabulation Data Validation

Document Submission in Various Formats –

Significance Testing Data Entry
PDF, HTML, XML, Word, Power Point

Posting and Charting of Tabulated Data Conjoint Analysis Data Tabulation

Data Cleaning

Statistical Analysis using

Open Ended Coding
SPSS

12
 Weighting

Survey data analysis

Data can do beautiful things, but turning your survey results into clear, compelling
analysis isn’t always a straightforward task. We’ve collected our tips for survey analysis
along with a beginner’s guide to survey data and analysis tools.

Survey analysis is the process of turning the raw material of your survey data into insights and
answers you can use to improve things for your business. It’s an essential part of doing survey-
based research.

There are a huge number of survey data analysis methods available, from simple cross-tabulation,
where data from your survey responses is arranged into rows and columns that make it easier to
understand, to statistical methods for survey data analysis which tell you things you could never
work out on your own, such as whether the results you’re seeing have statistical significance.

How to Choose the Right Survey Method

Prior to launching a survey, it’s imperative to consider what resources are available to
you to conduct the surveys, your goals and objectives, the sample and population size
and many other factors to consider:

 Time: Typically in-person surveys require more time than online surveys. Consider how much
time you have available and how soon you need results.
 Money: Consider what your budget is and how much it will cost to conduct your survey of
choice.
 Audience: Consider your audience demographic. Online surveys are typically more difficult for
an older population. Consider what type of surveys yield a higher response rate for your
audience.
 Technology Capabilities: Consider whether you have the proper technology that can handle
your data analytics.
 Research Expertise: Consider whether you have the proper research knowledge such as clearly
defined problems, objectives, a target audience, proper data tracking capabilities etc.
 Responses: Consider whether you have strategies in place to combat slow response or no
response rates.

The Most Common Types of Surveys

There are so many survey types, but we’ll explore 7 types of surveys in depth since
these are the most common and beneficial for most companies.

Online Surveys

Online surveys are one of the most popular types of surveys and for good reason.
Online surveys are easy to create, disseminate, and gather responses—challenges

13
many other survey types face. Using an online survey platform, your company can
create a survey easily and quickly—often with a wide range of question types—and
send it out to customers within seconds. Your customers can then answer the survey at
their convenience and send their responses back to you instantaneously.

In addition, online surveys are beneficial for companies that have a wide reach and
customers around the world. Online surveys are accessible by anyone anywhere with
only the click of a button. You can reach a wide audience with this type of survey.

But like any survey type, online surveys do have their cons. Online surveys could be
accessed anywhere, but the survey recipient has to have internet access. There are
parts of the globe and some households that don’t have internet access (or easy access
to public internet) that could be valuable contributors to your survey. In addition, online
surveys can provide a wealth of information, but they fall short in some depth. There’s
no organic way to ask follow up questions or to probe deeper into an answer when you
want more insights.

Paper Surveys

Paper surveys are sometimes looked at as old fashioned, but they do still reach an
audience that can’t access many other types of surveys. Many audiences—like senior
citizens or those without internet access—are far more likely to respond or far more
comfortable with paper surveys than online surveys. Plus paper surveys are often
easier to read for many since they’re printed in large fonts with black text on white
paper.

But one major con of paper surveys is that they can be one of the pricier types of
surveys. Paper surveys have to be printed off, sometimes in large quantities, to reach
your audience. That can add up if you’re trying to reach a large amount of people. In
addition, paper surveys aren’t the most environmentally friendly way to survey.

Mail Surveys

Mail surveys are one of the most traditional types of surveys. While some people may
look at them as old fashioned, there’s a reason this method has been around for so
long: it feels authentic and offers you access to an audience that may be less inclined to
respond online. Those demographics who are less likely to answer an online email
survey are often familiar with mail surveys and will be willing to respond. In addition,
mail surveys also have a wide geographic reach since most places are accessible by
post.

When it comes to mail surveys, there are a couple of things to plan for in advance. First,
creating the surveys can often be simple, and you do only need to print them off to
distribute. But you should keep in mind that your response rate will often be lower if you
don’t pre-paid return envelopes for the participants. One drawback of mail surveys is
that they require some extra work on the part of the survey participants—something that

14
often lowers response rates. You’ll want to reduce that level of work as much as
possible by providing return envelopes that simply need to be dropped in a mailbox.

That brings up the second con of mail surveys: they can cost your company a little bit
more. The cost of printing, envelopes, and postage can add up—but could be worth it
for the responses you receive. Second, when it comes to response rate, less is more.
Mail surveys have a better completion rate if they’re short. Try to keep your survey to
about a page. Not only will that increase your chances of completed responses, short
surveys also reduce printing and postage costs.

Telephone Surveys

Telephone surveys are another more old fashioned method of surveying that has been
around since landlines. Now that nearly everyone carries a phone with them or has one
at phone, telephone surveys are alive and well in surveying methods. For a telephone
survey, a real live interviewer will ask a series of questions and record responses. With
modern technology, it’s easy for interviewers to insert responses directly into a
computer system to chart and track data.

In addition, telephone surveys let interviewers add a personal touch to the conversation
that isn’t possible for online, mail, or paper surveys. The chance to ask follow-up
questions can give the interviewers a chance to probe for deeper insights and
emotions.

Telephone surveys still face a few cons. These surveys may be cost effective, but they
should be kept short. Fifteen minutes is the cap for an effective phone survey. In
addition, phone surveys run the risk of looking like a telemarketing call or being ignored.
With the saturation of telemarketing calls, many cell phone users don’t answer calls
from unknown numbers. Some even block all unknown numbers. While telephone
surveys can be effective, your team of interviewers will first have to be ignored a decent
amount.

In-Person Surveys

While paper and online surveys lack a personal touch, in-person surveys excel at that.
With these surveys, interviewers are able to directly interact with a respondent and to
ask follow up questions that really delve into the nuances and intricacies of a response.
These surveys provide accuracy and profound insights. In addition, in-person interviews
are a great way to collect responses from audiences with low literacy—something that
may hinder groups from participating in other methods.

While the insights gleaned from these surveys can be immensely valuable, there are
still some cons to consider. In-person surveys can be time consuming. Your team will
need to prepare a venue, schedule interviews, and spend the time actually performing
the interviews. It’s a very hands-on process the entire time. In addition, in-person
interviews run the risk of interviewer bias or inexperience. The interviewer is crucial to

15
getting usable insights, so you will need to spend the time and money training or hiring
one. Even a well-trained interviewer can still lead to interviewer bias skewing your
information.

Panel Surveys

Panel sampling is when you select a group of people to survey repeatedly over a period
of time. Your respondents will be randomly selected from your target audience, and
they’ll respond to several surveys over a determined period of time. Longitudinal studies
are a great example of panel surveying. These surveys can provide long-term insights
and show in-depth understanding of a target audience. In addition, your company could
use a third-party research company to perform a large bulk of the research and to
ensure it’s done at a high level of quality.

But panel surveying takes time and preparation to make them effective and usable. In
addition, some participants may choose to stop responding to surveys part way through
the study, leaving you with gaps. Your company may need to provide incentives to
encourage participants to follow through with the entire surveying process.

Focus Group Surveys

Focus groups are a variation of in-person surveying methods that includes a panel of 6–
10 people that represent the target population. All of these people will answer questions
and hold a discussion, moderated by a professional who attempts to keep everything on
track and as unbiased as possible. These surveys can reveal personal attitudes and
perceptions for a sampling of your entire target population—something that can be very
valuable for market research.

The cons of focus groups, though, include the costs of preparing a moderator and
participants. It can take time to select just the right panel of participants to make sure
you are accurately exploring your target audience. You may also need to pay for the
travel expenses and time of an effective moderator. Oftentimes you may also need to
survey your focus group participants a second time to gather quantitative data, and that
takes additional time and preparation.

Benefits of Using Surveys

If surveys take time, preparation, and work, why do companies keep turning them for
market research? Aren’t there easier ways? Well, there may be, but there are few ways
that can give you such insight into your target populations and how your brand is
performing. Surveys provide strategic benefits based on customer feedback that can
help improve CX and improve brand loyalty.

16
 Here are some other key benefits of using surveys:

 Large sampling size. Surveying can reach many customers in many locations or with many
limitations. With a variety of surveys, you can reach people all over the world, people who don’t
use the internet, and people who can’t normally participate in other marketing research
methods.
 More cost-effective. Even the pricier types of surveying are fairly cost effective when
compared to the valuable insights you can gain.
 Reliable and usable information. Surveys can give you information and insights that are a
reliable reflection of your audience, especially of how they feel toward many parts of your brand.
In addition, this information can actually be usable (with the right analysis) to improve your
brand, customer experience, and, ultimately, ROI.
 Improve the brand experience. Your customers deserve a brand they trust and have a
positive experience with. Using surveys, you can get real feedback from your customers
and improve your brand experience in meaningful ways.

Quantitative data vs. qualitative data

What’s the difference between qualitative data and quantitative data?

 Quantitative data, aka numerical data, involves numerical values and quantities. An example of
quantitative data would be the number of times a customer has visited a location, the temperature of
a city or the scores achieved in an NPS survey.
 Qualitative data is information that isn’t numerical. It may be verbal or visual, or consist of spoken
audio or video. It’s more likely to be descriptive or subjective, although it doesn’t have to be.
Qualitative data highlights the “why” behind the what.

Closed-ended questions
These are questions with a limited range of responses. They could be a ‘yes’ or ‘no’ question such
as ‘do you live in Portland, OR?’. Closed-ended questions can also take the form of multiple-choice,
ranking, or drop-down menu items. Respondents can’t qualify their choice between the options or
explain why they chose which one they did.

This type of question produces structured data that is easy to sort, code and quantify since the
responses will fit into a limited number of ‘buckets’. However, its simplicity means you lose out on
some of the finer details that respondents could have provided.

Natural language data (open-ended questions)

Answers written in the respondent’s own words are also a form of survey data. This type of response
is usually given in open field (text box) question formats. Questions might begin with ‘how,’ ‘why,’
‘describe…’ or other conversational phrases that encourage the respondent to open up.

17
This type of data, known as unstructured data, is rich in information. It typically requires advanced
tools such as Natural Language Processing and sentiment analysis to extract the full value from how
the respondents answered, because of its complexity and volume.

Categorical (nominal) data

This kind of data exists in categories that have no hierarchical relationship to each other. No item
is treated as being more or less, better or worse, than the others. Examples would be primary colors
(red v. blue), genders (male v female) or brand names (Chrysler v Mitsubishi).

Multiple choice questions often produce this kind of data (though not always).

Ordinal data
Unlike categorical data, ordinal data has an intrinsic rank that relates to quantity or quality, such as
degrees of preference, or how strongly someone agrees or disagrees with a statement.

Scalar data
Like ordinal data, scalar data deals with quantity and quality on a relative basis, with some items
ranking above others. What makes it different is that it uses an established scale, such as age
(expressed as a number), test scores (out of 100), or time (in days, hours, minutes etc.)

You might get this kind of data from a drop-down or sliding scale question format, among others.The
type of data you receive affects the kind of survey results analysis you’ll be doing, so it’s very
important to consider the type of survey data you will end up with when you’re writing your survey
questions and designing survey flows.

Steps to analyze your survey data

Here’s an overview of how you can analyze survey data, identify trends and hopefully draw
meaningful conclusions from your research.

1. Review your research questions

Research questions are the underlying questions your survey seeks to answer. Research questions
are not the same as the questions in your questionnaire, although they may cover similar ground.

It’s important to review your research questions before you analyze your survey data to determine if
it aligns with what you want to accomplish and find out from your data.

2. Cross-tabulate your data

18
Cross-tabulation is a valuable step in sifting through your data and uncovering its meaning. When
you cross-tabulate, you’re breaking out your data according to the sub-groups within your research
population or your sample, and comparing the relationship between one variable and another. The
table you produce will give you an overall picture of how responses vary among your subgroups.

Target the survey questions that best address your research question. For example, if you want to
know how many people would be interested in buying from you in the future, cross-tabulating the
data will help you see whether some groups were more likely than others to want to return. This
gives you an idea of where to focus your efforts when improving your product design or
your customer experience.

Cross-tabulation works best for categorical data and other types of structured data. You can cross-
tabulate your data in multiple ways across different questions and sub-groups using survey analysis
software. Be aware, though, that slicing and dicing your data very finely will give you a smaller
sample size, which then affects the reliability of your results.

1. Review and investigate your results

Put your results in context – how have things changed since the last time you researched these
kinds of questions? Do your findings tie in to changes in your market or other research done within
your company?

Look at how different demographics within your sample or research population have answered, and
compare your findings to other data on these groups. For example, does your survey analysis tell
you something about why a certain group is purchasing less, or more? Does the data tell you
anything about how well your company is meeting strategic goals, such as changing brand
perceptions or appealing to a younger market?

Look at quantitative measures too. Which questions were answered the most? Which ones
produced the most polarized responses? Were there any questions with very skewed data? This
could be a clue to issues with survey design.

2. Use statistical analysis to check your findings

Statistics give you certainty (or as close to it as you can get) about the results of your
survey. Statistical tools like T-test, regression and ANOVA help you make sure that the results
you’re seeing have statistical significance and aren’t just there by chance.

Statistical tools can also help you determine which aspects of your data are most important, and
what kinds of relationships – if any – they have with one another.

Benchmarking your survey data

19
One of the most powerful aspects of survey data analysis is its ability to build on itself. By repeating
market research surveys at different points in time, you can not only use it to uncover insights from
your results, but to strengthen those insights over time.

Using consistent types of data and methods of analysis means you can use your initial results as
a benchmark for future research. What’s changed year-on-year? Has your survey data followed a
steady rise, performed a sudden leap or fallen incrementally? Over time, all these questions become
answerable when you listen regularly and analyze your data consistently.

Maintaining your question and data types and your data analysis methods means you achieve a like-
for-like measurement of results over time. And if you collect data consistently enough to see patterns
and processes emerging, you can use these to make predictions about future events and outcomes.

Another benefit of data analysis over time is that you can compare your results with other people’s,
provided you are using the same measurements and metrics. A classic example is NPS (Net
Promoter Score), which has become a standard measurement of customer experience that
companies typically track over time.

How to present survey results

Most data isn’t very friendly to the human eye or brain in its raw form. Survey data analysis helps
you turn your data into something that’s accessible, intuitive, and even interesting to a wide range of
people.

1. Make it visual

You can present data in a visual form, such as a chart or graph, or put it into a tabular form so it’s
easy for people to see the relationships between variables in your crosstab analysis. Choose a
graphic format that best suits your data type and clearly shows the results to the untrained eye.
There are plenty of options, including linear graphs, bar graphs, Venn diagrams, word clouds and pie
charts. If time and budget allows, you can create an infographic or animation.

2. Keep language human

You can express discoveries in plain language, for example, in phrases like “customers in the USA
consistently preferred potato chips to corn chips.” Adding direct quotes from your natural language
data (provided respondents have consented to this) can add immediacy and illustrate your points.

3. Tell the story of your research

Another approach is to express data using the power of storytelling, using a beginning-middle-end or
situation-crisis-resolution structure to talk about how trends have emerged or challenges have been
overcome. This helps people understand the context of your research and why you did it the way
you did.

4. Include your insights

20
As well as presenting your data in terms of numbers and proportions, always be sure to share the
insights it has produced too. Insights come when you apply knowledge and ideas to the data in the
survey, which means they’re often more striking and easier to grasp than the data by itself. Insights
may take the form of a recommended action, or examine how two different data points are
connected.

Common mistakes in analyzing data and Solution

1. Being too quick to interpret survey results

It’s easy to get carried away when the data seems to show the results you were expecting or
confirms a hypothesis you started with. This is why it’s so important to use statistics to make sure
your survey report is statistically significant, i.e. based on reality, not a coincidence. Remember that
a skewed or coincidental result becomes more likely with a smaller sample size.

2. Treating correlation like causation

You may have heard the phrase “correlation is not causation” before. It’s well-known for a reason:
mistaking a link between two independent variables as a causal relationship between them is a
common pitfall in research. Results can correlate without one having a direct effect on the other.

An example is when there is another common variable involved that isn’t measured and acts as a
kind of missing link between the correlated variables. Sales of sunscreen might go up in line with the
number of ice-creams sold at the beach, but it’s not because there’s something about ice-cream that
makes people more vulnerable to getting sunburned. It’s because a third variable – sunshine –
affects both sunscreen use and ice-cream sales.

3. Missing the nuances in qualitative natural language data

Human language is complex, and analyzing survey data in the form of speech or text isn’t as
straightforward as mapping vocabulary items to positive or negative codes. The latest AI solutions
go further, uncovering meaning, emotion and intent within human language.

Trusting your rich qualitative data to an AI’s interpretation means relying on the software’s ability to
understand language in the way a human would, taking into account things like context and
conversational dynamics. If you’re investing in software to analyze natural language data in your
surveys, make sure it’s capable of sentiment analysis that uses machine learning to get a deeper
understanding of what survey respondents are trying to tell you.

21
 Tools for survey analysis
If you’re planning to run an ongoing data insights program (and we recommend that you do), it’s
important to have tools on hand that make it easy and efficient to perform your research and extract
valuable insights from the results.

It’s even better if those tools help you to share your findings with the right people, at the right time, in
a format that works for them. Here are a few attributes to look for in a survey analysis software
platform.

 Easy to use (for non-experts)

Look for software that demands minimal training or expertise, and you’ll save time and effort while
maximizing the number of people who can pitch in on your experience management program. User-
friendly drag-and-drop interfaces, straightforward menus, and automated data analysis are all worth
looking out for.
 Works on any platform
Don’t restrict your team to a single place where software is located on a few terminals. Instead,
choose a cloud-based platform that’s optimized for mobile, desktop, tablet and more.
 Integrates with your existing setup
Stand-alone analysis tools create additional work you shouldn’t have to do. Why export, convert,
paste and print out when you can use a software tool that plugs straight into your existing systems
via API?
 Incorporates statistical analysis
Choose a system that gives you the tools to not just process and present your data, but refine your
survey results using statistical tools that generate deep insights and future predictions with just a few
clicks.
 Comes with first-class support
The best survey data tool is one that scales with you and adapts to your goals and growth. A large
part of that is having an expert team on call to answer questions, propose bespoke solutions, and
help you get the most out of the service you’ve paid for.

How to analyze survey data

Reporting on survey results will prove the value of your work. Learn more about statistical analysis
types or jump into an analysis type below to see our favorite tools of the trade:

 Conjoint Analysis
 T-Tests
 CrossTab Analysis
 Cluster Analysis
 Factor Analysis
 Analysis of Variance (ANOVA)

22