CAI pictures, video and sound

From QPSMR 2004.2, telephone interviewing can only be done with QPSMR Telin. This new program is similar to QPSMR Input but extends the program to handle specific CATI requirements including sample management, quota control and appointments.  QPSMR Telin requires a licence and will be shipped to all QPSMR CATI users.

New empty projects will be set to ASCII (character) data; if you would prefer to use a  binary project simply set the [Binary data] switch in "Global project settings".

Undefined switches

The [Empty numbers as undefined] and [Filtered numbers as undefined] switches are now automatically set ON, so any integer, weight, date, or time, questions or variables that are filtered past or skipped (left blank) in data entry will be “undefined” and they can be referred to in filter or variable definitions as U.

Format option FSL – “Footnote to show Significance Levels used” is now set on as default (to comply with quality standards). This means that a footnote will automatically be displayed on all tables where significance testing is applied (using letters or * markers)

Format LBF1 – “Lines before footnote set to 1”, is also now applied as default, so 1 blank line will be left at the bottom of the table(s) before the footnote is printed

From the “Question worksheet” the [Preselect responses] and [Auto select responses] buttons may be used to automatically select responses to the question.  These selected responses may be those chosen at a previous question, or they may be variable responses that hold derived information from other entries.  The [Preselect responses] setting means that the question will be displayed during data entry/interview; questions with [Auto select responses] set are not shown.  These new settings can be useful to speed data entry with long lists of responses, in product tests where codes need to be entered several times, and they are especially useful for CAI projects when a number of responses need to be selected at random. 

Carry forward texts for easier CAI research!

It is now possible to carry forward texts to responses, interviewer instructions and hold filters (previously this facility was only available for use in question texts).  Also, it is now possible to carry forward from a multi coded question where the chosen responses will be displayed each separated by a + (plus sign).

CATI quota checking

For non-complex CATI quota control you no longer need to include ZZZQCC, ZZZQCP and ZZZQCF entries in your projects. If you do not include ZZZQCC entries, all “quota” entries (that is those with the [Quota] check box selected at the “Question worksheet”) will be checked.  If ZZZQCP is not included in your project, when the last quota entry is passed the interview will be counted as “in progress”.  If ZZZQCF is not included, when the ZZZEND entry is passed the interview will be counted as finished.

CATI supervisor reports

Two new CATI supervisor reports have been included in 2004.   [Supervisor] [List of serials] [List of serial records (current)…] produces a view of a generated CSV file with only one mention of each serial number (the current). [Supervisor] [Results of calls…] can be used to see the number and outcome (last 10 tries) for your CATI project.

Two new switches are available from the “Variable worksheet”.  Setting the [Ignore case when testing characters in response definitions] check box on means that QPSMR will not distinguish between upper and lower case characters when they are used in response definitions for the variable.  For integer or weight variables it is possible to set the [Treat undefined (U) as zero in arithmetic] check box, so that any numbers which are undefined (that is they were skipped during data entry and the [Empty numbers as undefined] switch is set on, or filtered past and the [Filter numbers as undefined] switch is set on) will be treated as 0 in any arithmetic definitions. 

New tables CSV file selections

Two new “Export CSV file” switches have been included so you have more control over the layout of the Tables.csv file that is automatically created when your tables are run.  The [Round figures as tables] selection ensures that figures and percentages output to the Tables CSV file use the same rounding as used in your QPSMR tables (for example, to 1 decimal place).  If you are using significance markers in your analysis the [Extra column for markers] selection will cause the markers in the Tables.csv file to be displayed alongside the figures, in a separate column, rather than below the figures and percentages.

Minimum records to show

In order to preserve confidentiality of respondents for projects with a limited sample, for example, specialist medical research, where it might otherwise be possible to identify respondents, it is possible to set a “minimum records to show” value.

When set a “minimum records to show” value will be retained and used for any QPSMR Reflect files (or CL files) generated from the project.

Reject category responses now stored in data

Responses with [Reject category] applied to them, use a data location, will be shown in data input (as data is entered) and the response number is now stored in the data file. 

TIP: If you use the [Reject category] setting, it is important that both your data entry and data verification is done using the same version of QPSMR, as older versions did not store reject response numbers and the program will flag this as a difference under verification.

Other 2004 features …

When quantity weighting tables an unweighted row (without the weight applied) will be included.

Range checks on numeric entries honoured in consistency checking.

Automatic back up files

Whenever you save your work using QPSMR version 2003, a back up file is automatically created and a copy of the master file (.QDF, .QTF, .QRF, .ASC .CBA) is saved into a sub-folder called Archive (also created automatically).  The name of the file is appended with an underscore followed by 4 digits.  For example, the first copy of a questions definitions file (.QDF) saved will be numbered _0001.QDF, the second _0002.QDF and so on.

Once 20 copies of the respective file have been saved by QPSMR into the Archive folder, the program will display a message stating that there are a number of copies of the file in the Archive folder, suggesting you may wish to delete the lowest numbered (oldest) copies.  We recommend that you do this, otherwise your project folders will use a lot of hard disk space.

KFILE folder

This release does not create internal files when edits, data lists and tables are run, instead temporary files are used.  These temporary files are stored in the folder set by your computer system, typically in a folder on the C drive.  This means that disk usage is reduced, and if the project resides on a network, traffic will be reduced if the temporary folder is on the local C drive.  If you use QPSMR export facilities, a KFILE sub-folder will still be created to store internal files (which will be copied by the program to the folder you specify during the export process).  The KFILE sub-folder may therefore be deleted at a later date.

Output files

The extensions used for files output from QPSMR have changed, most output files are now TXT files (with the main file name extended) for example, Project.TAB will now become Project Tables.TXT.  This means they can easily be opened simply by double clicking. 

[File] [List files …] removed

The menu selection [File] [List files…] from the “Project” window, is no longer available as it is not needed with the file name changes described above.

Access to score values

CSV table improvements

The tables CSV output now includes formats MED (Median), MOD (Mode) and ILE (Quantitles) if set on your tables.

New format FSL puts a footnote on your table(s) to show the significance levels used for testing.  This should be used on all tables which have markers to show significance levels, to comply with quality standards.

New format LBF allows you to insert blank lines before any footnotes at the bottom of tables.

Dialog boxes moved

Dialog boxes displayed by the program are now positioned towards the right of your screen (away from the left hand margin) so you can easily see entry and table names.  This is especially useful if you need to refer to several existing entries in a new filter or variable and prefer to type syntax!

New MakeQDF.sbs file

A file called MakeQDF.sbs is now shipped with the QPSMR software and is stored In the Examples folder automatically created by the program when it is installed.  This SPSS script file allows users with SPSS to generate a QPSMR question definitions file (.QDF) and corresponding ASCII data file (.ASC) from an SPSS SAV file.

New file names

As with the menu system (described above) many output files now have the extension .TXT, which enables them to be accessed easily using your editor.  To distinguish the output files, the file name itself is appended with an additional word to describe the contents.  For example, "Project.stp" used to output the tables to "Project.tab", but will now output tables to "Project Tables.txt".

New RCP OLDNAMES

A new RCP  OLDNAMES, retains the old names for output files for users who have batch runs which rely on the old output file naming conventions.

New RCP TIDY

RCP TIDY, should be used with most set-ups as it suppresses the generation of the KFILE files, thus saving disk space and reducing network traffic where project files are held centrally.  Note that TIDY, cannot be used if the set-up is needed for export or input into another set-up.

New RCP NOADDINI

This RCP NOADDINI, prevents CL from adding this set-up to the list of recently accessed set-ups in the IniSTP.txt file.

New RCP MINSHOW

The RCP MINSHOW=1, prevents data lists from this set-up and also from exported Reflect files from this set-up.  It also prevents further exports from output Reflect files.

New functions %WMIN and %WMAX

These functions are similar to %MIN and %MAX but set an mvar (or svar) with the locations of the lowest or highest numbers.  For example -

dm $which=%wmax($i1,$i2,$i3),

will set the true the responses that correspond to the highest values.

If $i1=3, $i2=1 and $i3=3, then responses will be set dm $which=t,f,t,

Format UZA (undefined as zero in arithmetic)

A new logical format UZA causes the value U (undefined) to be treated as zero in arithmetic definitions (DI and DW).

For example under F=UZA the definition -

    di $sum=$i1+$i2+$i3,

will give the sum of the three integers even if one or more of them are U.  If they are all U the answer will be 0.

A new logical format FSL causes special %FSL (see below) to be output to show the significance levels used for marking cells.

New special text %FSL shows the significance levels used for marking cells as a footnote when format FSL is set.  This text can be changed provided certain rules are followed.

A new special text %UND can be changed to set the text used for U (undefined) rows in VDT tables in place of "---".

CSV2 output has been improved with MED, MOD and ILE output.

Menus and toolbars have been altered to allow direct access to output files.

Far East character (CHARSET=DOUBLE) output has been enhanced.

CATI Appointments and other features

This release includes an appointment system that incorporates ...

New check boxes to set [Sample points] and [Quotas] from the "Question worksheet" (so these are automatically used with Sample index files and quota files are built)

New ZZZ entries to store appointment information

A "list of appointments made" report and a "serial history" report  available for CATI supervisors

Appointments are shown at the "Start of interview" dialog box ...

Interviewers can make vague or fixed appointments, for themselves, other named interviewers or for any interviewer by making selections from the "Appointment" dialog box. 

Interviewers also decide the appointment priority and set an acceptable call time delay; they request a "respondent history" where dates and times the interview has previously been tried, and the result of the call (for example, no reply, engaged) are shown.  Comments about the interview or the respondents may also be typed as required.

Filter application and removal

QPSMR 2002.2 has a 2 new selections available from the “Project” window, [Project] [Apply filter to select entries] and [Stop applying filter to selected entries].  Simply select a filter and a number of other entries to which the filter should be applied or removed and select the appropriate option.

Variable responses

A [Select all] button has been added which allows you to select all of the responses in a multi or single variable.  A new [Special] dialog box allows selected responses to reversed (swapped around) and sorted alphabetically on the response texts (first word, second word or third word).

Question responses

Question responses can now be copied from Word, WordPad, Excel or other programs - each new line in the copied information becomes a new response to your question.   

Data location list improvements

[View] [Data location list] now includes Serial/Card and Job data locations.  This option now also reports any faulty data locations.  

Data location setting

With version 2002.2 when defining your questionnaires you can now specify that the program should allocate a data location at the “Beginning” of the project, “Normal” (as listed) or “End” to use a location from the end of the list.  This feature will normally be used for ZZZ and associated entries.  For example, for CAPI and CATI projects, to reduce the size of data files all ZZZ entries and any questions following ZZZEND can be marked [Beginning].  For paper surveys, it is possible to mark ZZZ entries with [End], so the answers to ZZZ entries will be stored at the end of the data file, regardless of their position in the project (list). 

The empty page at the beginning of the printable tables file (.TAB) is no longer produced.

Routing instructions moved

The routing instructions (Question routing, Text routing as well as Ineligible and Close for responses) have been moved so they now appear on the “Question worksheet” and the “Response handling for question” dialog box respectively, thus making them more accessible.

QPSMR version 2002.2 includes a “first” version of an XML Triple-S (.SSS) import.  As this is a preliminary version, the imported file will need to be thoroughly checked after import and be revised as required.

Effective sample size

New format UNE causes the unweighted figures in a respondent weighted table to be replaced by the sum of the squared weights.  At print time the “unweighted” figures are replaced by the “effective sample sizes” (the weighted base squared divided by the sum of the squared weights).

These replacement figures will always be less than the unweighted figures and they will be used in all statistical calculations.  The more the individual weights vary from 1.0 the lower the “effective sample size”.

By producing a separate additional table with formats UNE and NPTR (PTR set off) applied the weighting effect on the total row figures can be seen. 

Character questions in analysis

Up to 245 characters may now be shown when analysing character questions (in previous versions only the first 20 characters were displayed).  Please note that only identical responses (including upper and lowercase characters) will be amalgamated in your analysis, for example “Red” and “red” would not be combined.  If you use character questions as rows in your analysis you may need to increase the row label width from the default 24 characters (by changing format option RLW) in order to display all of the character text.

QPSMR version 2002.1 has a new facility available under [Project] [Select>] [Choose…] which allows you to search for entry names.  You can search for entries by referring to a letter or a number (or a set of letters and/or numbers which appear one after another) at the beginning, middle or end of the name.

The QPSMR CATI [Supervisor] [Quota control] [Edit quota file] selection now includes extra [Plus 100] and [Minus 100] buttons. 

Update / verify in Input

When updating information, once a record is saved, the program will now show the serial number for the next questionnaire in the data file at the “Which serial number” dialog box.

When “Verify questionnaire” is requested QPSMR will automatically show the first questionnaire in the data file that has not yet been verified (the entry for which ZZZUSV is empty).

Data reports

When [Data] [Run reports] is used to produce a holecount or a frequency count, the report now includes the data file name at the top of the first page, so it possible to see at a glance which data file was used to generate the information. 

Selected entries for topline tables

When requesting [Tables] [Create topline tables file] any entries that are already selected (highlighted) at the “Project” window will be pre-selected as row entries, so it is now an even quicker process to produce topline tables containing only certain entries. 

Tables window

The “Tables” window now shows (stats only) alongside relevant tables.

The “tables” included in the CSV file generated each time tables are run, or copied and pasted to your spreadsheet program, now look much more like your QPSMR tables!  For example, format options such as BRS (Blank Row Suppression), BCS (Blank Column Suppression) and CON (Continuation table) are all honoured in the export. 

Significance markers

All significance markers produced by formats SIG (Significance tests on distributions) and TTF (t tests / F tests) will now take effect even if your break runs onto a number of pages.  In previous versions only columns that appear on the same page could be tested. 

Format CHI2 causes the CHI square figure to be shown even if more than 20% of cells have an “expected” figure of less than 5.  

Exporting enhancements