Additional Variables in the Data Set

There are additional variables before (left) and after (right) of your question's variables. This chapter will shortly describe their meanings.

Note: Some variables must explicitly be enabled before starting the download.

Note: For privacy reasons, recording data from the user's browser (browser, referer, IP address, etc.) needs to be to activated before collecting data.

• CASE Unique number of the respondent. As a rule, the number corresponds to the order in which the respondents opened the questionnaire (also see the PHP function caseNumber()).
  • Whenever someone opens the questionnaire, a new number is assigned – if the person does not click Next or opens the questionnaire again right away, an empty record results. By default, it is deleted automatically. In addition, case numbers are also assigned when testing the questionnaire. As a consequence, the numbering usually does not start at zero and has gaps (e.g., 123, 125, 130, 131, 132, …).
  • Optionally, you can enable the option “Keep records of empty interviews” in the Project Settings in the tab Privacy. This prevents the deletion of empty cases and thus the gaps in the numbering.
  • Case numbers are also assigned for tests during questionnaire development. It is therefore common that the number does not start at 1. To keep the case numbers unique per survey project, the counter cannot be reset – this would not bring any advantages in the analysis either.
• SERIAL If the questionnaire was opened with a personalized link (access code), the participant ID appears here (also see the PHP function caseSerial()).
• REF If a reference was passed in the link to the questionnaire (The Questionnaire's URL), the reference text is stored here (also see the PHP function reference()).
• QUESTNNR Contains the ID of the questionnaire that was filled out. The ID is specified when composing the questionnaire. An ID “del:<number>” means that the questionnaire with which the interview was conducted has been deleted.
• MODE Indicates whether the questionnaire was started in a special mode:
  • “interview” means that the questionnaire was opened normally via the link
  • “pretest” marks questionnaires from the pretest (with display of the question IDs and feedback option)
  • “orgtest” marks questionnaires from the pretest that were started in the original layout
  • “admin” means that the questionnaire was started in the admin preview ()
  • “debug” marks interviews that were tested in debug mode ()
• LANGUAGE Language in which the interview was conducted. This variable is included in multilingual survey projects or if the option Download variables which have not been used in the questionnaire is active. If the language was changed during the interview, the language used last is given.
• STARTED Indicates the point in time at which the respondent opened the questionnaire.
• MAILSENT In interviews that were started by means of a personalized mailing link, the time of sending the mailing is stored here. But only if the address entry uses the privacy mode “individual-related”. Otherwise, it may be possible to use the function panelData().

The following variables are found at the end of the dataset.

• LASTDATA Point in time when the respondent most recently clicked “Next” and thereby updated the record. The interval between STARTED and LASTDATA may deviate from the sum of the response times, because the processing times of the web server are fully included here.
• STATUS If an interview status was set for the interview, automatically or manually by means of setStatus(), it is given here.
  • complete – The case counts as a completely filled out questionnaire.
    As of version 3.8.00, the status “complete” is assigned automatically when the last page is reached, if no other status has been set (more on this in the manual Quota Control) and if at least some of the questions were answered (threshold according to the setting under Compose Questionnaire in the tab “Access”).
  • quota count – The case counts as a completely filled out questionnaire, but no redirect has been triggered yet.
  • screenout – A screenout was triggered because the person does not belong to the population.
  • quota full – The person was rejected because the quota for the controlled characteristics had already been reached.
  • quality fail – Control questions or other indicators led to an exclusion due to insufficient data quality.
• FINISHED Did the respondent reach the “last page” (1) or not (0).
• LASTPAGE Indicates which page of the questionnaire was answered last and submitted with “Next”. The number corresponds to the page numbering in the questionnaire (Compose Questionnaire).
• MAXPAGE The last-placed page in the questionnaire that was answered. In contrast to LASTPAGE, this is not the page answered last, but the answered page with the highest number. The value corresponds to that of LASTPAGE, unless (a) the respondent used the back button (e.g., to look at the first page once more at the end) or (b) jumps “forward” are programmed in the questionnaire.

• TIMEnnn The variables TIME001 etc. store how long (in seconds) a participant has spent on a page in the questionnaire. The time from loading the page to submitting it by clicking “Next” is specified.
  • These times are only imprecise, as they also include loading times. The typical inaccuracy is in the range of 1-2 seconds, but can be higher in individual cases (e.g. with an unstable internet connection). More precise measurements are possible using JavaScript: LatencyTimer
  • The number in the name of the variable refers to the page number under “Compose Questionnaire”, TIME007 always refers to page 07 - regardless of where the page was displayed during the interview (e.g. because pages were skipped or presented in a different order due to rotation).
  • If a participant sees a page several times (e.g. by using the back button in the questionnaire), the times are added together.
  • If the loopPage() or loopToPage() functions are used, then the variable indicates the cumulative dwell time for all repetitions of the page.
  • If a respondent leaves the browser window with page 5 of the questionnaire open overnight and fills in page 5 the next day and then clicks on “Next”, a dwell time >20,000 seconds (several hours) may be shown in TIME005.
  • If several pages of the questionnaire are displayed at the same time (e.g. because one page shows no content or because goToPage() is used), the dwell time is saved for the first page that shows content. Example: Page 8 only contains a PHP code with setPageOrder(), followed by page 9 with a question and a goToPage() to page 10, where a question is also shown, then the participant sees the questionnaire pages 8+9+10 all together on one page (i.e. 2 questions one below the other) and the dwell time is saved in TIME009.
• TIME_SUM The sum of dwell times (in seconds) after correction for breaks. If the participant suspends the interview and returns later, this seems like he or she stayed on the page for a long time (hours or even days). Such times are replaced by the other participants' page median. Dwell times are identified as break if
  • they are longer than 2 hours or
  • they exceed the page's median dwell time by more than 3 inter quartile ranges (IQR) divided by 1.34 (equals more than 3 standard deviations in a normally distributed sample)
  • Note: Which times are identified as outliers depends on the downloaded or displayed (partial) dataset. The values under Collected Data → View Data therefore do not always match the values in the download.
• TIME_RSI An index that indicates how much faster a participant has completed the questionnaire than the typical participant (median) has done. Values above 1 identify faster respondents, values below 1 slower respondents. Details see below.

Note: The parameters TIME_SUM and TIME_RSI only contain a value if the downloaded data set contains at least 10 records for the respective questionnaire (filter_criteria). The more records the download contains, the more accurate the values for TIME_SUM and TIME_RSI will be, because the distribution of response times in the sample is used to clean outliers or to normalize them.

Note: The response times are only included in the data set if the option Dwell time has been checked in the variables selection when downloading the data (Download Data). This option is checked by default.

Note: Processing times are recorded automatically. To deactivate the recording, please uncheck the option in Survey Project → Project Settings → tab Privacy → record time and duration during the survey.

Data quality in online surveys is usually quite good. Data cleaning, however, is necessary in mostly every survey. When using the option Variables selection → Download data quality parameters SoSci Survey provides variables to support data cleaning:

• MISSING The percentage of answers omitted by the participant (0 to 100). Only such questions and items are counted that have been shown to the participant – therefore someone dropping out early may have answered all questions (to this page, 0% missing). This variable is useful to identify participants that just viewed the questionnaired.
  • Please note that no click in a checkbox question (multiple selection) is a valid answer. Therefore even void cases may not reach 100%.
  • “Don't know” options are counted as valid answers as well.
  • When using text inputs, an invalid answer is counted, if the respondent types nothign (or spaces, only). Please remember this, when optionally asking for texts (e.g., when the respondent may leave the text field empty instead of writing a zero).
  • When using Free text inputs within a selection (single or multiple choice selection), a option's void text input (e.g., “Other: ___”) is only counted as invalid data, if the appropriate option in the selection was selected.
• MISSREL Percentage of missing answers weighted by the other participants answering behavior. Questions that are rarely answered (e.g., voluntary text questions) are mostly irrelevant for this value, questions that most participants have answered weight worse. The linear weighting factor for a question/item is the number of answers given to this question/item divided by how often the question/item has been asked.
  Note: This value may vary, depeding on the subset of data retreived.
• TIME_RSI This parameter is documented in detail in the article Too Fast, too Straight, too Weird (named “relative speed index”). Records with a value in the range of 2.0 and above should be considered critically. However, knowledge questions that the participant may need to research may distort the value (participants with good prior knowledge are faster).
• Q_VIEWER In the questionnaire you can enable the option “Option to view the questionnaire without answering mandatory questions”. When the function is active, this variable indicates whether a participant has checked the corresponding box (“I only want to view the questionnaire”).

The variables LASTPAGE and FINISHED can be used to determine whether a questionnaire has been completed in full (see above). The proportion of missing information (MISSREL) is a valuable indicator of the diligence of the participant or for data sets that originate from “just having a look”. Although the time invested in completing the questionnaire is not a direct indicator of data quality, very low response times (low TIME_SUM and high TIME_RSI) indicate that the questions were not even read.