DataLoader : New Features

DataLoader additional features guide 

Welcome to the DataLoader training tutorial. In this guide we will give you an overview of the more advanced DataLoader features, and why you might need them. Each section includes a short, hands-on video, showing the functionality in action.

Here is a checklist for getting started: 

• An account and login details to access DataLoader, which can be accessed here: https://platform.telmar.com/login 
• An SPSS file, the SPSS FILE requirements can be found here. 
• Remember that the maximum SAV file size that is supported

What’s New?  

• Dashboard usability 
• Advanced codebook editing functions 
• Importing a codebook 
• Adding weights to a survey 
• Applying a base filter to the survey

Improved dashboard functionality 

Making it easier to: 

• Sift through multiple uploaded surveys, search for a survey code, locate surveys by name, by person as well as verifying the status of the upload.

Quick tour: Dashboard functionality

Creating Net responses 

Making it easier to: 

• Combine multiple answers with an OR. For example: 
  • Men OR Women = All respondents 
  • Age 18 OR Age 19 OR Age 20 = Age 18-20 

Quick tour: NET responses 

Creating Advanced Net responses 

Making it easier to: 

• Combine multiple answers together to establish a total of similar or bulk answers. For example: 
  • Grouping ‘Strongly agree’ and ‘Tend to agree’ statements into ‘All Agree’ 
  • Creating heavy, medium and light users 
  • Grouping product usage together 

Quick tour: Advanced NET responses 

Creating answers 

Making it easier to: 

• Create your own answers. 
• Some surveys have an empty field or ‘no answer’, typically referring to the sum of all respondents. 
• With these questions, you are able to define the answers. Common answers that can be created are: 
  • Age 
  • Gender 
  • Answer scales (typically 0-10) 

Quick tour: Creating answers

Creating volumetrics 

Making it easier to: 

• Code totals or averages; specify the decimal points 
• Control what midpoints are used to calculate volumetric results (see example below). 

Volumetric coding options include: 

• MEAN: used to calculate the average against the total sample (including respondents who were not asked or who did not answer the question).
• MEANZ: used to calculate the average for those that answered the question, but excluding respondents who answered None/Zero 
• MEDIAN: the value at which 50% of respondents are above and 50% are below that value 
• MEDIANZ: the value at which 50% of respondents are above and 50% are below that value, excluding respondents who answered None/Zero 

Midpoint examples: 

• To create volumetrics for monthly frequencies, choose the variable equal to 1 and build on that, in this example, ‘at least once a month’ 
  • Every day - midpoint: 30 
  • At least once a week - midpoint: 4.2 
  • At least once a month - midpoint: 1 
  • Several times a week - midpoint: 10 
  • 2-3 times a month - midpoint: 2.5 
  • Once every 2-3 months - midpoint: 0.3 
  • A Less often - midpoint: 0.1 

Midpoints for Income group answers: 

• Up to $7000 - midpoint 3500 
• $14001 - $21000 - midpoint 17500 
• $28001 - $34000 - midpoint 31500 
• $41001 - $48000 - midpoint 44500 
• $55001 - $62000 - midpoint 58500 

Quick tour: Volumetrics

Inverting questions or answers 

Making it easier to: 

• Group answers together. The inversion function will swap answers and questions 
• All questions under the category must have the same answers in order to use the inversion function 
• For example: A question in the survey may be focused on the frequency of social media use. Instead of manually coding the answers in SurveyTime, we can invert the questions to group the answers in DataLoader. 

Let’s dive into an example: 

Quick tour: Inversion

Importing the codebook 

Making it easier to: 

• Re-purpose a codebook that has already been created for another survey where the surveys are the same or similar (part of a series of surveys) 
• Import a previously created custom codebook into the new or latest wave of data provided that the variable name as remained the same. 
• This is very useful for tracker type data or proprietary datasets that have little changes between waves 

Quick tour: Importing the codebook 

Adding weights 

Making it easier to: 

• Adjust the weighting of the survey to match the population profile in order to get a representative sample (each respondent in the sample represents a number of individuals) 
• Correct sample imbalance. 

Updating population units 

Making it easier to: 

• Change the population unit for reporting purposes in SurveyTime/Explore. 
• The options are: 
  • 1 
  • 10 
  • 100 
  • 1,000 
  • 10,000 

Quick tour: Weights and Population Units 

Adding weights 

Making it easier to: 

• Select the weight variable from the source file 
• You also have the following options: 
  • Specify whether this is the default weight or not 
  • Apply a weight adjustment factor method (divide or multiply) 
  • Set the weight adjustment factor value 

Quick tour: Adding weights 

Adding base filters 

Making it easier to: 

• Exclude respondents that have not answered a question (All Blank Data). This provides a more accurate representation of the data in a survey analysis. 

Adding custom base filters 

Making it easier to: 

• Exclude specific variables that may not be required in the cross tab analysis. For example, excluding respondents aged 18+ from the question, Frequency of drinking alcohol. This will remove all respondents under the age of 18 that may have answered the question. 

Quick tour: Adding base filters 

Importing Base Filters 

Making it easier to: 

• Re-purpose previously created base filters. This feature is very useful for wave type surveys where the questions remain the same or are similar between different survey waves. 
• Only base filter codes that exist in both datasets can be imported. The application will red flag codes that are not present in the new data upload. 

Quick tour: Importing base filters

Stability flags 

Making it easier to: 

• Flag the results in the crosstab, where the sample size is below a specified number 
• One star * indicates that the data is relatively unstable. 
• Two stars ** indicates that the data is highly unstable 
• The default numbers ranges are : 
  • One star * = 100 respondents 
  • Two stars ** = 50 respondents 
  • Shown as 100/50 
• If no stability warning is preferred the levels can be set to 0/0 

Quick tour: Adding stability flags