|
| |
 
Listed below
are some commonly asked QPSMR questions ....
For more detailed information, please refer to
the on-screen help within the QPSMR program itself.
If you have a question you would like to email
to the QPSMR Help-line select QPSMR.
|
Q: How many responses can I use for each single or multi coded question?
A:
Up to 9999 responses are allowed for each single or multi coded entry.
|
|
Q: How many entries can I define in my questionnaire?
A: Using QPSMR you may define 30,000 entries per survey (questionnaire).
|
|
Q: How can I skip certain entries in data input?
A: By applying a filter to every question which is not asked to all
respondents (see on-screen help section Filter Overview for more
details).
|
|
Q: How can I test the filters I have applied to my questionnaire, before
entering real data?
A: QPSMR has a "test data entry" facility which allows you to enter
test data for a questionnaire, without storing the data in a data file and
without setting the "Data files exist switch" on. This facility
enables you to work through your questionnaire entering data and checking that
the "Type" of questions are correct (single/multi response etc) and
that you have allocated sufficient data locations to store integer and/or weight
questions. See on-screen help section Test input data for more details.
|
|
Q: How do I add a new question and still keep all
of the data I have already entered?
A: You should use
[View] [Data locations list] to see the last data location used. Next, set the [Data files exist]
switch OFF and [Locations fixed switch] ON under "Global questionnaire
settings". It will then be possible to include a new question - but you
must manually specify a new, unique, data location for the new question and all
of it’s responses, QPSMR will not stop you setting the same data location twice
(although data input will not be permitted if a duplicate data location is
present).
|
|
Q: How can I see a summary of the data collected so far?
A: To
create topline tables select [Tables] [Create topline tables file…].
|
|
Q: How many tables can I run at once?
A: Up to 30,000 tables can be stored (and run) in a single QPSMR table
definitions file (.QTF) any number of .QTF files can be used in the same
project.
|
|
Q: How do I include more than 1 entry across the top (as the columns) of my
tables?
A: By making a variable. You can use the [Block insert…] selection from the
"Response handling for variable" dialog box to include copies of whole
questions and then modify the responses if required.
|
|
Q: How can I include column headers on tables?
A: By using a \ (backslash). When the responses are used as the columns of a
table, the response text up until the \ will be used as a header, centered over
the remaining text and the following response texts (up to the next response
text with a backslash). When the responses are used for the rows of a table,
backslash generates a new line. For example, you might use the following:
Variable Response texts
Response 1 text: Sex\Male
Response 2 text: Female
Response 3 text: Age\Under 30
Response 4 text: 30 and over
Table column headers
Sex
Age
Table 1 Male Female Under 30 30 and over
Note: Format options UCH and URH can be used to underline column and row
headers.
|
|
Q: How can I print more than 1 table on the same page?
A: Use format option LBT0 or LBT1, to leave no lines, or 1 line between each
printed table.
|
|
Q: I have defined a print style - why don’t my tables use all of the page,
when I print them?
A: Selecting [Data] [Global report settings...] or [Tables] [Set table
defaults...] or clicking the [Set defaults...] button from the "Global
table settings" dialog box, will position you at the "Global table and
report settings" dialog box.
Here you can select a pre-defined print style, by clicking the [Print
style...] button . We recommend that
you use the [Print style...] button to select a pre-defined style to control the
page size rather than altering the width or height of the page yourself, in this
way you can be sure that QPSMR will automatically use all of the available page
space for your reports and tables. This will prevent information, when there is
space available on the current page, overflowing to the next.
IMPORTANT: If you modify a print style, in order for the changes to take
effect, you must always select the style, using the [Print style...] button from
the "Global reports and tables settings" dialog box - even if the
style is already displayed.
|
|
Q: Why don’t the columns of my tables line up correctly when I print them?
A: If you are using a proportional font (such as Arial or Times Roman) to
print your tables, you should apply format options CHG2 and CLG2 globally,
before running your tables, to ensure that the columns line up correctly.
|
|
Q: How do I differentiate between valid zero responses and blanks (skipped
answers) on my tables - especially in mean scores?
A: From the "Global table and report settings" dialog box you can
request [Empty numbers as undefined] which means that any integer or weight
questions which are empty (left blank in data input) will be set to
"undefined" meaning that they will not be included in any table
calculations (they will be counted as undefined rather than zero).
|
|
Q: How can I change the "Mean score" text in a mean score summary
table?
A: From the "Global table and report settings" dialog box,
selecting the box [Averages use question text] enables you to use the question
texts as the row labels in place of the standard text "mean score",
when producing a "statistics only" table or a "mean score
summary" table. It is also possible to include a CL syntax table %AVG='text',
see on-screen help section CL syntax tables.
|
|
Q: How can I produce an index (table of contents) for my
tables?
A: A "Table of Contents" print file is automatically created each
time you run a set of tables. The table of contents print file has the same name
as the tables file followed by Contents.TXT. This can be printed from
the "Project" window by selecting [File] [Print a file..].
Note we recommend that a special print style .PRS file is created for
printing table of contents files, using an [Extra font] with the settings [Start at
line] 1, [Finish at line] 99; [Begin at text block] 1, [Finish at text block] 99
and setting the [Left] radio button.
|
|
Q: When I have finished a project, which files should I back off
?
A: The question definitions file (.QDF), the data file (.CBA or .ASC) and the
table definitions file (.QTF). If you have used "respondent weighting"
in your analysis you should also keep the weighting targets file (.WGT), and if
you have created a specialised Command Language set-up file, you should also
back off the CL set-up script file (.STP). All other files can be re-created by QPSMR.
|
|
Q: My data processing bureau does not use QPSMR, what type
of data file should they send to me?
TIP: The QPSMR
Input program
is available for distribution to in house data entry staff, home workers and to
other companies who provide a data capture service.
Data from a variety of other programs may be analysed with QPSMR. Many types of
data file may be used directly with QPSMR, but others may need modification before
they can be processed.
It is possible to
import .SSS (triple S) data into QPSMR; the .SSS file type is the market
research “standard” for the interchange of data. To import a .SSS file, from the
“Project” window select [File] [Import a file >] [SSS import…]. You should then
choose the .SSS file you wish to import from the “Select SSS file for import”
dialog box, and choose a name for the question definitions file (.QDF) to be
generated at the “Select new project name” dialog box. QPSMR will then produce a
question definitions file (.QDF) to use with the ASCII data file (.ASC).
Most types of column binary data file can be used with QPSMR, although they may need to be "copied and converted" into a .CBA
data file using the [Data] [Convert raw data…] selection before use.
Both types of ASCII data can be used with QPSMR, fixed format (where the file
is held in blocks of 80 columns) and flat ASCII (data locations begin at 1 and
continue until the record is complete).
Another type of data
file which is commonly used with QPSMR is a .CSV file (Comma separated variable)
usually output from spreadsheet or database programs.
|
|
Q: Even though I have created a weighting targets file (.WGT) for respondent
weighting, my tables are not being weighted when they are run - why?
This is probably because when some word processors save TEXT ONLY files they
add their own .TXT extension, so the resulting file is actually called
xxx.WGT.TXT (where xxx is the name of the weighting targets file). If so, QPSMR will not be able to locate a .WGT file and so your tables will not be weighted.
To check the name of your weighting targets file, use the Windows Explorer to
list the file, and select [View] [Folder options…] [View] and ensure that the
[Hide file extensions for known file types] check box is NOT selected. In this
way the entire weighting file name will be displayed. If it does have the
"added" extension .TXT you can then rename it and your tables will be
weighted when they are run.
|
|
Q: When I open a data file in QPSMR, I see a screen which lists a number of
duplicate cards and duplicate serial numbers – what do I do?
If ever you see a message in QPSMR which refers to duplicate card or duplicate
serial numbers, it is very important not to ignore it, because the program
suspects that there may be something wrong with your data!
If you see this message and you are SURE that NO duplicate questionnaires are
present in the data file, it is possible to use the [Sort and compact] facility
to remove the "duplicates" from the file. However, as the QPSMR menu system will include only the "last
mention" of any duplicated serial number when you view your data file, we
recommend that you take a note of the duplicate serial numbers from the
"View" screen message. It is then possible to use the [Change serial]
facility to allocate a new unique serial number to the "last
mentioned" duplicate serial number which you can access by updating the
appropriate serial number. If you then exit from the "Data" window and
then open the data file again, unless the serial number has been duplicated more
than once, it will no longer appear in the "View" window message. It
is then possible to update the original serial number followed by the "last
mentioned" serial (to which you allocated a new unique number) and
establish if there is in fact a duplicated record. If so, and the records are
identical, you can then delete one of the questionnaires.
|
|
