Omnibus: Weekly tasks

Introduction

A new Danish CATI omnibus is to be configured every Monday. Test links should be delivered to the client no later than 1400 DK time. The production start is usually 1600 or 1700 DK time.

Due to possible late changes in last weeks bus, the bus for next week should not be configured before the last bus is completed. (Usually Sunday evening.)

The ideal situation is keeping the monday configuration unchanged for the entire week. However, from time to time we get new projects with short notice, and sometimes these are added to the bus during the week.

Tasks

Step 1 - Preparation

Get the list of projects/questionnaires to be included in the new week: Bus specification

• In order to remove a project from the BUS, remove its id from the new bus' property Sub questionnaire ids
• In order to add a project to the BUS, make sure it has the common format stated in Catglobe Omnibus System 2009, add its id to the new bus' property Sub questionnaire ids]

Prepare all sub questionnaires, get their primary keys by using Actions - Information in questionnaire resource list.

Step 2 - Create a new bus questionnaire

Create a new questionnaire using Omnibus questionnaire resource template

• Name: 9031cat - BUS DK [YYYYWW]
• General settings:
  • Layout: Layout resources\CGTemplate (questionnaire layout template): CATI
  • Path: CATINET\Internal Projects\9031cat - 2009 Omnibus Setup
  • Questionnaire template: 9031cat - 2009 Omnibus

• Properties
  • Demographics questionnaire id: primary key of demographics questionnaire
  • Opinion poll questionnaire id: primary key of opinion poll questionnaire
  • Catinet Panel Recruiting: primary key of Catinet Panel questionnaire
  • Sub survey ids: primary keys of sub questionnaires: Ritzau, CEM Economy, CEM Lifestyle... (order is important)
  • Other Panel Recruitings: primary keys of other panel questionnaires: TV2, CityMail, ...

Step 3 - Import quotas

Quota template file can be found at \\catproc\share\Catglobe\Projects\Catinet\CAT\9031cat - 2009 Omnibus Setup
This quota file contains only the quotas for questionnaires that usually are included in the bus.

• Remove quotas of sub questionnaires that are not supposed to be included in the new week
• Add quotas of sub questionnaires that are to be included in the new week (make sure to use the same names as in D_Quotas question's sub question texts)

Step 4 - Test your work!

Preview the bus questionnaire and go through it to see if there are any errors. Make sure that all questionnaires are in Danish version

Step 5 - Prepare samples

We need to make sure that there are enough samples for CATI department to call to achieve the current week's target.

Samples come from last week's unused respondents and fresh respondents who have not answered any bus questionnaires.

Last week's samples

We need to move last week's respondents who have not been used yet: not started or or just been through the very first question in the questionnaire.

1. Open group BUS - DK - Last week's sample

• Rule=Partly
  • Variable 1: change the questionnaire to last weeks questionnaire
  • Variable 2: change the questionnaire to last weeks questionnaire
    • Select the question D_QAS_Status, check Like below and check Not started
  • Variable 3: change the questionnaire to last weeks questionnaire
  • Save and close the rule
• Rule=Not started
  • Variable 1, 2: change the questionnaire to last weeks questionnaire 
  • Save and close the rule
• Rebuild the group

2. Rebuild 2 groups below

• BUS - DK - Last week's sample - Normal phones
• BUS - DK - Last week's sample - Mobile phones

3. Start moving respondents using Move respondents wizard

 

• Source questionnaire: last week's questionnaire (it is recommended to set as default setting after the first use)
• Destination questionnaire: current week's questionnaire (set as default setting)
• Group: repeat the process with 9 groups below:
  • BUS - DK - Last week's sample - Ålborg - Mobile phones
  • BUS - DK - Last week's sample - Ålborg - Normal phones
  • BUS - DK - Last week's sample - Århus - Mobile phones
  • BUS - DK - Last week's sample - Århus - Normal phones
  • BUS - DK - Last week's sample - KBH/FRB - Mobile phones
  • BUS - DK - Last week's sample - KBH/FRB - Normal phones
  • BUS - DK - Last week's sample - Odense - Mobile phones
  • BUS - DK - Last week's sample - Odense - Normal phones
  • BUS - DK - Last week's sample - Region Hovedstaden - Mobile phones
  • BUS - DK - Last week's sample - Region Hovedstaden- Normal phones
  • BUS - DK - Last week's sample - Region Midtjylland - Mobile phones
  • BUS - DK - Last week's sample - Region Midtjylland - Normal phones
  • BUS - DK - Last week's sample - Region Nordjylland - Mobile phones
  • BUS - DK - Last week's sample - Region Nordjylland - Normal phones
  • BUS - DK - Last week's sample - Region Sjælland - Mobile phones
  • BUS - DK - Last week's sample - Region Sjælland - Normal phones
  • BUS - DK - Last week's sample - Region Syddanmark - Mobile phones
  • BUS - DK - Last week's sample - Region Syddanmark - Normal phones
• Answer sheet status: Unchanged
• Answer sheet phase: Unchanged

Fresh samples

Open the fieldwork managment interface and study the Sample tab for the current bus questionnaire.

The sum of the figures in the Main list column is required to be at least 15 times the gross target of the survey. (For example, if the omnibus has a gross target of 1.000, the sum of the main list column should be no less than 15.000.)

If the sample size is insufficient, fresh samples must be added.

Check if the sample is representative

Calculate percentage of 5 groups of regions based on Main list:

• Region Ålborg & region Nordjylland
• Region Århus & region Midtjylland
• Region KBH/FRB & region Hovedstaden
• Region Odense & region Syddanmark
• Region Sjælland
  

After calculating we compare the result to the standard file "Calculate group targets.xls" from this Path:
\\catproc\share\Catglobe\Projects\Catinet\CAT\9031cat - 2009 Omnibus Setup

If the percentage result of our main list is similar to the file "Calculate group targets.xls", do nothing; or else, we have to adjust the number of respondents for our regions.

1. Build the following 18 groups

• BUS - DK - Ålborg - Mobile phones
• BUS - DK - Ålborg - Normal phones
• BUS - DK - Århus - Mobile phones
• BUS - DK - Århus - Normal phones
• BUS - DK - KBH/FRB - Mobile phones
• BUS - DK - KBH/FRB - Normal phones
• BUS - DK - Odense - Mobile phones
• BUS - DK - Odense - Normal phones
• BUS - DK - Region Hovedstaden - Mobile phones
• BUS - DK - Region Hovedstaden - Normal phones
• BUS - DK - Region Midtjylland - Mobile phones
• BUS - DK - Region Midtjylland - Normal phones
• BUS - DK - Region Nordjylland - Mobile phones
• BUS - DK - Region Nordjylland - Normal phones
• BUS - DK - Region Sjælland - Mobile phones
• BUS - DK - Region Sjælland - Normal phones
• BUS - DK - Region Syddanmark - Mobile phones
• BUS - DK - Region Syddanmark - Normal phones

2. Create a new group, name it with: BUS - DK - Week yyyyww

Add all of 18 groups above to the new group and rebuild it.

3. Add the newly created group to group DK - Tlfnumre - brugte

Rebuild this group

4. Add 18 groups above to the new bus questionnaire as sample rules, and rebuild them

• Open the bus questionnaire
• Open Samples tab - Sample information - Edit total sample

• When adding groups to the questionnaire, always leave empty as sample size

Step 6 - Verify your work!

Preview the questionnaire one more time
Open Fieldwork Management tool to check if we have enough samples for this week

Step 7 - Send test links to the fieldwork manager

Generate the test links and send them to izam.mehanic@norstat.dk, cc to vnconsultants@catglobe.com

Step 8 - Inform CEM Economy team about the bus

For the time forward, Catglobe Economy data is always collected when we're running the omnibus, and should be upgraded the following Monday (interviews are completed Saturday evening). Therefore, we need to inform CEM Econonmy team (Luan for now) about the new bus, so that he will schedule some time updating the diagrams on monday the week after.