Chapter 9

• Company


  Research Now, a leader in digital research data
• Partner


  Carat, a global media agency
• Client


  N Brown, a British digital retailer
• Objective


  To measure the impact of N Brown's digital advertising across different media channels on brand awareness and perception
• Challenge


  • N Brown, traditionally focused on above-the-line media, needed to justify its increased digital spending
  • Carat required a comprehensive view of how various advertising channels affected consumer awareness and perception
• Tool Used


  Research Now's Adimension® Cross-Media measurement solution
• Methodology


  Persistent first-party cookies tracked ad exposure across media channels
• Data Collection


  Linked multiple ad exposures to create an 'exposed' group and a control group
• Assessment


  Custom-designed surveys evaluated campaign effectiveness across different channels
• This study demonstrated the effectiveness of using advanced cross-media measurement tools to assess and refine advertising strategies in real-time
• Population


  The entire group under study as defined by research objectives
• Census


  A complete accounting of the entire population
• Sample


  A subset of the population that represents the entire group
• Sample Unit


  The basic level of investigation within the sample
• Sample Frame


  A master source list of sample units from which the sample is drawn
• Sample Frame Error


  Errors that occur when the sample frame does not perfectly represent the population
• Sampling Error


  Errors that occur due to the use of a sample to estimate characteristics of the entire population
• Define the Population


  Clearly specify who or what is being studied
• Determine the Sample Frame

  Identify a source list that encompasses the population
• Identify and Minimize Sample Frame Errors

  Ensure the sample frame accurately reflects the population

  
• Select the Sample


  Choose a representative subset from the sample frame

  
• Conduct a Census or Sample Based Study


  Depending on the research scope, either account for every unit in the population or select a sample for study
• Address Sampling Errors


  Recognize and mitigate errors stemming from the sampling method and sample size
• For practical application, like in marketing research, it's often impractical to conduct a census, making sampling a necessary and efficient technique. Ensuring the sample accurately reflects the population is critical for reliable research outcomes
• Probability Sampling


  Members of the population have a known chance of being selected
• Probability Sampling Methods


  • Simple Random Sampling
  • Systematic Sampling
  • Cluster Sampling
  • Stratified Sampling

  
• Nonprobability Sampling


  The probability of any member being selected is unknown
• Nonprobability Sampling Characteristics


  • Subject to human intervention and biases
  • Often used when exact probability calculations are impractical
• Probability methods are used when the research requires quantifiable and generalizable results
• Nonprobability methods are more common in exploratory research or when the research conditions do not allow for a structured random selection process
• Simple Random Sampling


  1. Assign Unique Numbers

  2. Generate Random Numbers

  3. Select Corresponding Members

  4. Repeat Until Sample is Complete

  5. Skip Duplicates
• Systematic Sampling


  1. Complete List Required

  2. Calculate Skip Interval

  3. Random Starting Point

  4. Select at Intervals
• Systematic Sampling Advantages


  More efficient than simple random sampling due to fewer random draws and systematic coverage of the list
• Systematic Sampling Disadvantages


  Vulnerable to errors if the list is outdated or incomplete (e.g., unlisted numbers in telephone directories)
• How to Take a Systematic Sample

  1. Obtain a Population List

  2. Calculate the Skip Interval

  3. Select a Random Starting Point

  4. Apply the Skip Interval

  5. Address Any Skipped Entries
• Stratified Sampling


  Analyzes each stratum (subgroup) within the population separately, ensuring that all variations within the population are adequately represented
• Stratified Sampling Benefit 1

  Explicit Analysis of Each Stratum
• Stratified Sampling Benefit 2


  Weighted Mean Calculation

  
• By using stratified sampling, researchers can ensure that their samples are not only representative of the population in terms of size but also in terms of the internal diversity and complexity of the population. This leads to more precise and actionable insights, particularly in heterogeneous populations where different segments may exhibit distinct characteristics

  
• Convenience Sampling


  Selection is based on ease of access. Examples include intercepting people in high-traffic areas like malls