Understanding survey raw data

1. Introduction

If you want to upload survey interview data from other systems into Survalyzer or create customised dashboards or analyse raw data in a third party tool a good understanding of Survalyzer question types and their raw survey data is essential. The following document explains the raw data structure of the seven most common question types.

Survalyzer Know How Prerequisites

you need to have a good understanding of some basic concepts of Survalyzer.

A: How to build a survey

B: How to distribute a survey

C: How to use the analyse a survey

2. Variable name and variable values

Survalyzer questionnaires consist of a set of question types. The following table shows the question types we will take a closer
look

1. Open question
2. Star Scale
3. Single Response
4. Multiple Response
5. Matrix
6. NPS

Each question type has its own specific data property, which are explained below. The two most important concepts are variable
names and variable values. These two concepts are described in the following chapter.

2.1 Variable names and variable values

Let us consider a single-response question type. First, we show the respondent’s view, and then on the right we show the relevant settings for this question type with three choices.

Respondent View

Setting view in Survalyzer

hints about 3 red numbers

1. Indicates the question type. The question type defines wether is it a single-variable/single-column (see chapter 3.1, 3.2, 3.3, , 3.6 or a multi variable/multi column question (see chapter 3.4, 3.5). A single-variable question means there is only one column in the raw data export and multi variable means there are multiple columns per one question type.
2. The variable (or variable name) is the name of the variable. In MS Excel this is the column title.
3. Variable value is the value assigned to an option, For example agree has the value 1, neither nor has the value 2 and disagree has the value 3.

Hint: Do not confuse ‘Variable suffix’ settings with ‘Value’ settings. Multi variable question type usually have a setting Variable suffix to define the ending of a variable name.

Value setting

It defines a distinct value in the raw data. If the value is 1, the corresponding variable stores the value 1 if agree is selected.

Variable suffix setting

Defines the last part of the variable name. if Variable suffix = 1, the corresponding variable name gets _1 at the end, e.g. Q1 becomes Q1_1 (variable suffix can be number of letter)

2.2 Missings

Variables can also have missing values and its important to understand that missing empty value is not the same a deliberate choice of “Not applicable”. The link explains about different type of missings and its export format. see download raw data, section Missings depending on output format

3. Question types and their raw data representation

3.1 Open question

Respondent view Open question

VariableName: Open
Values: any text value is allowed

Raw data view – open question

Hints about row 1 to 3

• In Interview 1 the value is “everything..”
• In Interview 2 the value is #N/A (or #N/V in
  german excel) as respondent selected N/A option
• Interview 3 is empty as user skipped this question

3.2 Star Scale

Respondent view star scale

VariableName: StarScale
Values:
1:1 star
2:2 stars
3:3 stars
4:4 stars

Raw data view – StarScale

Hints about row 1 to 3

• In Interview 1 the value is 3 as respondent selected 3 stars
• In Interview 2 the value is #N/A (or #N/V in german excel) as respondent selected N/A option
• Interview 3 is empty as user skipped this question

3.3. Single Response

Respondent view Single Response

Variable name: SingleChoice
Values:
1:strongly agree
2:agree
3:neither agree nor disagree
4:disagree
5:strongly disagree

Raw data view – Single Response

Hints about row 1 to 3

• In Interview 1 the value is 2 as respondent selected agree
• Interview 2 and 3 are empty as user
  skipped this questions or cancelled the
  survey

3.4 Multiple Response

Respondent View Multiple Response

Variable names: MultiC_1, MultiC_2, MultiC_3, MultiC_3_text
Multiple Choice Questions
are a collection of binary
(0/1) variables. Value 1
stands for checked. The
text field is behind the third
choice therefore its infix is
_3_
Values of variables
MultiC_1, MultiC_2, MultiC_3:
{0:unchecked},{1:checked}

Raw data view – multiple response

• In Interview 1 the 2nd and 3rd option is checked additionally text ebanking is in the variable MultiC_3_text
• In Interview 2 all variables get N/A as respondent selected last option “None of the above”
• In Interview 3 all options get value 0 as user saw question but did not click any variable

3.5 Matrix

Respondent view Matrix

Variable names: myMatrix_1, myMatrix_2
Values of variables
1:very bad
2:rather bad
3:neither bad nor good
4:rather good
5:good

Raw data view – Matrix

• In Interview 1 the rather bad for Service 1 (value 2) and very bad for Service 2 was selected.
• Interview 2: N/A option (not visible) for both Services are selected
• Interview 3: respondent skipped answers.

3.6 NPS (Net Promoter Score)

Respondent view NPS

Variable Name: NPS
Values for Variables: 0 to 10
Values cannot be changes as NPS is a static method

Raw Data Representation NPS

• Interview 1 respondent selcted the first choice, 0 very unlikely
• Interview 2 respondent selected the last choice, 10 very likely
• Interview 3: respondent skipped answers.