Survey dataSurvey data is queried using Eighty20's FP system. This system allows you to do simple cross tabulations of data. You select the row and column variables you want to see in your final table. You can also create a filter field to allow you to filter the data that appears in your table.
Selecting a surveyYou can access the survey data by moving your mouse to the "Survey Data" tab on the navigation bar at the top of the page. You will then see a drop down list of the available surveys.
Setting up your queryOnce you have selected the survey, you will see the page shown below.
- A - Click here to reset the query if you've made a mistake
- B - Click here to get the survey questionnaire
- C - This button allows you to select the row field of your table
- D - This button allows you to select the column field of your table
- E - This button allows you to create a filter for the data in the table
- F - Click here once your query has been set up to see your results
Selecting a row or column fieldIn order to select a row or column variable click on either the "Choose Row" or the "Choose Column" button. In the image below the "Choose Row" button has been selected. This brings up a list of category branches that can then be selected. Category branches will have a double arrow, shown below in red, and if clicked they will branch out into a more expanded selection of survey fields.
In the image below the "Demographics" category was selected. This brings up a list demographic survey fields. Survey fields will not have double arrows. These fields can be selected as the row/column in your table.
The same process can be used to select a column variable. Selecting a column variable is optional depending on whether you are interested in:
- A single variable - for example, seeing what types are households there are. In this case you would not need to select a column variable, you would only need to select "Housing Type" for the row variable.
- Or a cross tab - for example, looking at type of household by area. In this case you would need to select a column variable. You would need to select "Housing Type" for the row variable and "Area" for the column variable.
Setting up a filter
Choosing a filter to include in the analysis is optional. The filter allows you to retrieve data from a specific group only. For example, you could filter on a particular area or dwelling type.
The filter can be as simple or as complex as you like. You can simply select a survey field as a filter or you can use the logic buttons, shown in the image below, to create a more complex filter. If you wish to use the logic buttons simply click on them and they will be added to your filter. Remove logic elements and survey fields from your filter by double-clicking them.
The distinction between "And" and "Or" is important. If the filter is set as Gender:Female AND Marital Status:Single, the results will only include those observations for which the respondent was both female and single. If OR had been used instead, the results would include ALL female respondents (regardless of whether they are single or not) and ALL single respondents (regardless of whether they are female or not).
In the image below the filter of Gender:Female AND Marital Status:Single has been selected
Using your query resultsOnce you have chosen the row variable, and possibly column variable and filter you wish to query you can click on the "Get Data" button. This will bring up a results page like the one shown below.
- A - Click here to see the results in weighted numbers (this is the default setting)
- B - Click here to see the results in terms of the number of respondent observations
- C - This button creates a percentage row so that all items in the row add to 100
- D - This button creates a percentage column so that all items in the column add to 100
- E - This button takes you back to the previous query selection page
- F - Click on this button to download the questionnaire
- G - Click on this button to the results table to excel
Note that numbers are displayed in red when there are fewer than 40 observations as these numbers may not be statistically significant.