Skip to main content

FV Decipher Support

All the topics, resources needed for FV Decipher.

 
FocusVision Knowledge Base

Custom Report

  Requires Decipher Cloud

The custrep.txt file allows the creation of very customized reports, useful for monitoring response rates and other aspects of fielding for surveys, or even across multiple surveys. This report is viewable online and also available as a download into Excel.

For the custrep.txt file, basically each row has a little formula that gets evaluated. "Total" is the total number of responses (for qualified users, for all users you have to do all.total).

The custom report also has access to the actual question data for both real and virtual questions. Probably the best way to learn this file type is to look at actual custrep.txt files.

Here are some examples:

HTML version:

Excel/CSV version:

1: About the Syntax

. LRW Job Name: Service Evolution
. Some other text

@col . US N = 450
@col se Sweden N = 450

q5.r1.val Laptop Ownership
q5.r2.val Cell Phone Owner

. Business
virtq1 Flight Length: Short Range

So the general idea here is that the first word on a line can be: . ( a dot) to insert verbatim text, @col to define a column with subdirectory and title. As I see it, the N = XXX have to be hand specified.

Then follow the rows, the first part is an expression like in notify definitions, then the title.

(db.loaded - all.total) Total Drop-Offs

This is a a piece of Python code surrounded by () -- the () are not necessary if you just have something with no spaces that needs to be run.

The "db" is an special object that lets you dip into the dashboard data. The db.loaded is what's displayed as "Click Through" in the Dashboard report.

<>This is more complex:

((100.0*all.total)/(db.loaded or 1)) Incidence @avg |%.2f%%

We calculate the percentage so we multiply the all.total (i.e. all the respondents) by 100 and divide by load count (the or 1 ensures no division by zero).

The @avg bit makes the report calculate an AVERAGE rather than TOTAL for the total row.

The |%.2f%% bit specifies a format. The %.2f makes it display the floating point result with 2 digits of precision and the %% inserts a %.

#60*itime.getMean() Total Interview Time @avg |%d

is currently commented out, but if you label the interview time question as itime you will be able to get the average out, multiply it by 60 to display seconds and display it under each survey, with average in the total.

It is also possible to access the results of custom reports using @col syntax like:

@col survey (Standard report, "All" segment)
@col survey/report (This named report, "All" segment)
@col survey/report/segment (This named report, this named segment)

As usual, . in place of survey means current survey.

Anticipating overlapping segments I've added a @nototal option that will suppress the total row.

As mentioned before, both markers and db return survey-wide settings. You'll need virtual questions if you want to extract data for a specific segment.

To get the number of respondents who have seen any given page of a survey, similar to the data shown in the timing report, use:

pages.q9.seen

If needed in the future, we can extract the other information on the timing report with this syntax (e.g. pages.q9.completed / min / max etc.). This requires a completions.dat file that hasn't been reset

2: Percentages

If you need to display a percentage do it by using (top, bottom) rather than e.g. (100*top/bottom). This will make the total percentage work correctly.

3: Custom title cols

Per default, the row title you specify appears in a single column. If you specify @titlecol you can specify a custom or multiple custom column titles for those. If you specify multiple title cols, separate cell assignment using //. You must have exactly as many title cols in every row as @titlecol tags, for example:

@titlecol Region
@titlecol School
@head

(q1.r1.val) West // First Western School
(q1.r2.val) East // First Eastern School
  • Was this article helpful?