CASE 050

Welcome to CASE 050 - Principles of Panel Data Reporting

Panel helps our clients answer many important questions

Our goal today is to introduce you to the underlying Homescan principles and procedures necessary for successful application of CASE and other Nielsen Panel systems.

Today we will review 

• Components of Data Accuracy
• Statics
• Time Periods
• Projections
• Data Types
• Buyer vs. Shopper
• Data Sources

Please click the home button below to continue with this tutorial.

Please keep track of any questions you may have as you go through the session.  You will have an opportunity to submit them at the end of the session.

Components of Panel Accuracy

Please select true or false

• Sampling Error results from estimating a population by measuring a portion of that population
• Measurement Error is the largest component of non sampling error
• Nielsen can totally eliminate Bias from our panel
• We control for a standard error through our sample size
• Measurement error is the difference between the recorded response and the true value
• We very rarely discuss standard error with our clients
• Standard error measures the variability of an estimate due to sampling

Statics

Static -  a reporting criteria that is applied to households to ensure that only consistent reporters are included in the analysis.  This allows us to have confidence in the data. 

Nielsen Homescan Projections are carried out at the household level

Household sample must be

• Representative of the target population
• Good reporters to generate good projections

Independent Statics

An independent static is used to evaluate total market placed trends and should be used for most all panel analysis. 

• Each sample is projected to the total target population for each time period.
• Independent static will include the impact of household growth. 
  • You can have different households included in each period).
  • All households have to meet the static in each period.

Linked Statics

Linked Statics are used to trace a specific target population overtime.

• Not necessarily representative of the general market place, but is representative of how the same households behavior is changing over time.
• Generally used for very specific and focused business issues
• Samples have to meet the static in EACH time period
  • Linked statics are required for the following report models
    • Brand Shifting/Brand Measure
    • Sourcerer
    • New Lost Retained

Data Science Static Guidelines

Data Science Static Recommenation

* Monthly projections for the Trial and Repeat model should use a 1/2 static.  Reports that are aggregating 8 weeks of data should use a 2/2.

Please match the text to the description

• A reporting Criteria that is applied to households to ensure that only consistent reporters are included in the analysis
  Static
• Includes new households that are within our panel in the current time period but not in the previous time period
  Independent Static
• What we use when we want to follow households over time
  Linked Static
• Trial and Repeat Model should use what type of static?
  1/2 Static
• For a 12 month analysis what type of static is recommended?
  10/12

Defining Time Periods

In CASE there are always two different time frames that must be defined in the period dimension:

1. Data Period(s) - defines the time frame for Data
2. Static Period(s) - defines the time frame for included Households

These two time periods are usually - but not always the same.

 

Which two time frames are required

• Static & Data
• Sample & Bias
• Data & Measurement

Tell us what you know about time periods

There are 213 different time periods required in CASE.  A static time period is the time frame to determine included householdsweeks wantedteam norms while the data time period defines the time frame for datahouseholdsweeks wanted

Data Cycle

In CASE a data cycle definition is required as well for both the Data and Static time periods.​

• 4-4-5 data cycle is a 12 period data cycle for clients who follow a 12 period calendar.
• 4-4-4 data cycle is a 13 period data cycle for clients who follow a 13 period cycle.

Data cycles are determined by several considerations:

• Team norms - what the team and client has set as their expectations.
• The total number of weeks in the reporting period must be included in the data cycle.
• If you are trying to match a previously run report.
• Data cycles are usually consistent across data and static time frames.

How many periods are in a 4-4-5 Data Cycle?

• 12
• 13
• 10

How many periods in a 4-4-4 Data Cycle?

• 13
• 12
• 10

What are three things to consider when determining your report data cycle?

• Team norms, total weeks required, if you are trying to match a previously run report
• Team norms, data period, and static period
• 4-4-4, 4-4-5, and data period

Are data cycles usually consistent between static and data time frames?

• Yes
• No

Demographic Variables Used in Panel Projection

9 standard demographic variables used to project up to the market you are processing.  

• Household size
• Household income
• Household head age
• Female head of household age
• Male head of household age
• Household head occupation
• Presence of children
• Race
• Hispanic (Yes/No)*

Hispanic is not part of the race variable.  Hispanic is considered separately as an ethnicity.  If a household answers yes to Hispanic, their household information will be copied over to the Hispanic panel.

The annual panelist survey is fielded in October to capture information on these and other demographic variables.  The information is collected and loaded into CASE for the 1st week of the following year.

How many demographic variables are included in standard projection?

• 9
• 8
• it depends on what demographics you are interested in.

Is Hispanic included in the race variable?

• No
• Yes

Please select the 9 demographic variables used in standard household projection

• Household Size
• Household Income
• Age of Children
• Presence of Children
• Head of Household Age
• Head of Household Occupation
• Pet Ownership
• Female Head of Household Education
• Male Head of Household Education
• Primary Shopping Channel
• Race
• Hispanic (Yes/No)

There are different types of data types you can process within CASE

Show us what you know about different data types

There are 32Severaldata types you can process in CASE.  ShopperTransactionalRetail All Outletdata is made up of households who shopped in your specific retailer, but didn't necessarily by your item in that retailer.  TransactionalShopperRetailer All Outlet data reports transactions that occurred within a specific retailer.  Retailer All Out LetShopperTransactional data is volume for a specific item with at least one purchase of this item in a specific retailer along with all other purchases of the item outside the retailer.

Definition of Buyer and Shopper

Please select the questions that can be answered with either buyer or shopper facts.

Manufacturer and Retailers

What is a transaction, how is it different from a buyer?

What is a Buyer?

• 
• 
• 

What is a Shopper?

• 
• 
• 

What is a Transaction?

• 
• 
• 

Additional Resources CASE Resources

Congratulations on completing the CASE 050 online tutorial.   You are now ready to take the CASE 100 Foundation and Reports classroom course.    

If you have any questions regarding the content of the course, or need clarification on something, please submit them to nabuylearninganddevelopment@nielsen.com.

Please take a few minutes to save the attached documents to your computer in preparation for attending CASE 100, as these materials will not be part of your course material.   

CASE 050 Virtual Job Aid Principles of Data Reporting

CASE 050 Homescan Glossary of Terms

The Homescan Data Enabler files can be found on the US Data Inquiry iShare site.  Please click here for a link to the US Data Inquiry site, and select markets to find the most current Data Enabler file. 

To EXIT THE COURSE and log your completion, click the blue icon at the top right and then select "Finish course."