Questions and Answers - The Collected Data
Previous Topic  Next Topic 

Since questions form the major part of any survey, it is important to understand what happens to the answers - the collected data!

PocketSurvey is unique in that the data collected in the answers is automatically organised for you based on how you design a question. Essentially the organisation depends on what data destination you choose when you define a question.

Note

Users often ask about sections. Sections are very simple, they organise your questions into sections, and they don't actually collect data themselves. Think of sections as the overall structure of your survey.


Data Overview

The output from a survey is in the form of one or two comma separated data files (.CSV files). Typically you will import these data files into a database program such as Microsoft Access or a spreadsheet such as Microsoft Excel for later analysis.

When imported into a database program each data file form you survey becomes a database table, with each record within the file becoming a table row, and each field within a record being a table column.

Tip

If you don't understand database terminology of tables, row and columns you will probably need to get a database expert within your company to explain it to you.


Parents & Children

PocketSurvey lets you specify one of two data destinations for any individual question. The terms used for these two data destinations are 'parent' and the 'child'. However, you can choose the actual names for the parent and child data destinations. You could call the parent and child any of the following: 'father' and 'son', 'house' and 'rooms', or 'plot and 'house'.

The important thing to realise here is that there is a relationship between the data destinations. The technical term for this relationship is 'one-to-many', but it is easier to think of it as a parent having one or more children.

For a very simple survey you will probably only need to collect your data into a single data destination, the parent. For more complex surveys you need to use this 'one-to-many' or 'parent-child' relationship where you want to capture some overall data and also some repeating data. For example, when surveying a house property you may not only want to collect some overall property information but also collect some detailed information about each room in the house. (A house has one or more rooms -- in other words 'one-to-many'.)

Note

If you do not wish to store the answer to a question you select a data destination of 'none'. You would probably do this when asking questions that only jump to other questions.


Family Tree!

This one-to-many relationship enables PocketSurvey to model the requirements of most, if not all, surveys.

In fact PocketSurvey is even more powerful than this - you can even link to other surveys and extend this parent child structure ad infinitum like a family tree! Take the asset or facilities management situation and imagine a building having many rooms, each room has many desks, each desk has many pieces of equipment, and so on.

Parent Data Destination

The parent data file has records with a field for the Survey Identifier and a field for each answer. The optional first (header) row will have the field names that will help you when you import into a database package. This is shown schematically below.

Survey Identifier

Question 1

Question 2

Question 3

0001

No

No

Yes

0002

Maybe

No

Yes

0003

Yes

Yes

Yes


The file containing the parent data is always named as SurveyName#ParentDataName.csv and is found in the survey folder.

Child Data Destination

Questions with child data destinations generate a slightly different data structure than that of the parent data file. There will always be a field containing the Survey Identifier similar to that in the parent table, as this is necessary to maintain the 'one-to-many' relationships.

However, because we can have multiple occurrences (or instances) of child question answers we need some way of uniquely identifying each answer to each individual question. This is done by introducing extra fields to contain: Section number, the actual Question number, and Instance. This is shown schematically below.

Survey
Identifier

Question
Number

Instance

Question
Heading

Answer

0001

1.2.1

2

Question 1

Yes

0001

1.2.1

1

Question 1

Yes

0001

2.1.3

1

Question 3

No

0002

1.2.1

1

Question 1

Yes


The file containing the child data is always named as SurveyName#ChildDataName.csv and is found in the survey folder.

Answers to Common questions

Common questions generate additional fields in the child data file, shown in yellow and red below, and are often used to organise repeated questions. The field heading is the question prompt.

Thus for each common question there will be an extra column in the child data file. This is shown schematically below.

Survey Identifier

Question
Number

Instance

Question
Heading

Answer

Common
Question 1

Common
Question 2

0001

2.4

4

Plot No

34

Blue

Faulty

0001

2.4

3

Plot No

56

Blue

Faulty

0001

2.4

2

Plot No

36

Red

OK

0001

2.4

1

Plot No

54

Ye;;pw

Faulty

Another major use of common questions is to provide a costing mechanism.

Tip

You can view the complete database specification for a survey design by views in the database specification report from the Report | DB Specification menu option.


Data destination selection

The data destination selection on the questions screen is shown below. In the illustration the name of the parent data entity is 'Site Data' and the name of the child entity is 'Plot Data'.

HandHeld Systems Ltd ©