Preparing Survey Data For Import

A survey broadly consists of the name of the survey, the questions, and the answers for each respondent. Questions can be of different type:

  • Single choice between two (yes/no) or more options
  • Multiple choice
  • Open ended

Within the framework of ATLAS.ti these concepts are mapped as follows:

Survey ConceptATLAS.ti ConceptPrefix
Open-ended questionCode (and code comment):: to separate code and comment
Answer to an open ended questionContent of a quotation (data)(none)
Single Choice question 0/1Document group.
Single Choice question > 2 optionsDocument groups for each value:
Multiple Choice questionDocument groups for each value#

Based on specific prefixes that you add to your variable names, ATLAS.ti interprets the column headers and cells of the Excel™ table in different ways and turns them into documents, the content of documents, document groups, quotations, codes, comments and code groups. Sounds complicated, right? Not so --just follow along, it is actually very easy!

Data are imported case-based. This means each row of the Excel™ table that is imported from the online survey tool is transformed into a document.

In addition to the answers to open-ended questions, demographic information like age, profession, or age group, answers to single choice questions (yes/no, or offering more than two options) and answers to multiple choice questions can be imported. Within the framework of ATLAS.ti these are mapped in the form of document groups, one group per value.

Syntax

SyntaxResults in ATLAS.ti
!The name in the cells is used as document name
~This content is used as document comments
^The name is used in the author field
&The information is used as creation date for a document. Use the following format: ISO8601
<When added, this column is ignored. Use this to exclude variables you do not want to analyze in ATLAS.ti
:Each cell value is turned into a group. The column header is used as main group name. For example: If you have two values in the column :Gender (male and female) two document groups with the following names are created:Gender::male and Gender::female. The documents are sorted into those groups accordingly.
.A document group is only created from answers coded with 1 or yes. For example: If the survey contains a question: "Do you think that children bring happiness?" and the answer choices are yes and no (0/1), then only one group is created and respondents who answered the question with yes are sorted into that group.
#Document groups are created for each entry in a cell, which are separated by a comma. Use this option for questions that allow multiple responses For example: "Which of the following foods do you like? Choose as many that apply." Possible answers are: chocolate, cheese, ham, oranges, apples, potatoes or carrots. Respondent 1 selects: chocolate, cheese, apples. Respondent 2 selects: chocolate, ham, potatoes. When importing the data you get a document group for each of the possible response choices and each document is added to all document groups that apply. Assuming the column label is #Selected Foods, the document for respondent 1 will be added to the document groups: Selected Foods::chocolate, Selected Foods::cheese, and Selected Foods::apples.

All column headers without a prefix are interpreted as codes; the text in the cells will be the content for the documents to be analyzed. If the question is longer, and you do not want to use the full question as code name, then the full question content can be added as code comment. For example, when using the column header 'Question 1::Please write down reasons why you want to have children', Question 1 is used as code name; the text after the two colons is used as code label.

You can import tables in xls or xlsx format. In case you experience a problem, save the table in .csv format and try again.

If you import the same table repeatedly, rows with already existing documents are ignored. This way, you do not have to wait until the last respondent has filled out the questionnaire.

When importing the same data again with updated information, you can only import new cases (i.e. rows in Excel), but not new questions (i.e. columns in Excel).