Phase 1: Owning the Dataset
Argument Analysis Table
| Text from Passage |
Analysis |
| "A mattress company has two stores, one in City X and the other in City Z" |
- What it says: Company operates in two locations
- What it does: Sets up the comparison scenario
- Key connections: These are the two variables we'll be comparing
- Visualization: Store X <-> Store Z
|
| "The company has advertised equally in newspapers in both cities" |
- What it says: Newspaper advertising is the same in both locations
- What it does: Establishes a controlled variable
- Key connections: This is NOT the difference between cities
- Visualization: Newspaper ads: \(\mathrm{X = Z}\)
|
| "but has advertised twice as much on the radio in City Z as in City X" |
- What it says: Radio advertising differs between cities \(\mathrm{(Z = 2X)}\)
- What it does: Identifies the key variable difference
- Key connections: This IS the difference being tested
- Visualization: Radio ads: \(\mathrm{X < Z\ (Z = 2X)}\)
|
| "The two cities have similar populations and economies" |
- What it says: Other factors are controlled/similar
- What it does: Eliminates alternative explanations
- Key connections: Rules out demographic differences
- Visualization: Demographics: \(\mathrm{X ≈ Z}\)
|
| "and the sales at each store have been roughly equal" |
- What it says: Despite different radio advertising, sales are the same
- What it does: Provides the key observation
- Key connections: This is what leads to the consultant's conclusion
- Visualization: Sales: \(\mathrm{X = Z}\)
|
| "A consultant claims this shows that the radio advertising has not improved mattress sales" |
- What it says: Radio advertising is ineffective
- What it does: States the conclusion to be tested
- Key connections: Based on equal sales despite unequal radio advertising
- Visualization: \(\mathrm{More\ radio ≠ More\ sales}\)
|
Argument Structure
- Main conclusion: Radio advertising has not improved mattress sales
- Supporting evidence: City Z has twice as much radio advertising but equal sales
- Key assumption: The current advertising mix allows us to isolate radio's effect
- Logical flow: Different radio advertising → Same sales → Radio doesn't work
Phase 2: Question Analysis & Prethinking
Understanding What Each Part Asks
We need to select changes for:
- Part 1 (City X): A change that helps test the consultant's claim
- Part 2 (City Z): A change that helps test the consultant's claim
- Key insight: The two changes must work together to create a valid test
Question Type and Prethinking
This is asking for changes that would test a claim. To test whether radio advertising works, we need to:
- Isolate the effect of radio advertising
- Remove confounding variables
- Create a cleaner comparison between the cities
Specific Prethinking for Each Part
For both cities: The most direct test would be to eliminate the radio advertising entirely to see if sales are affected. This would:
- Directly test whether radio advertising has any positive effect on sales
- Provide a clear yes/no answer to the consultant's claim
- If sales remain the same, radio doesn't help; if sales decrease, radio does help
Phase 3: Answer Choice Evaluation
Evaluating Each Choice
"Double newspaper advertising"
- What it says: Increase newspaper ads to twice the current level
- For testing the claim: This adds more confounding variables rather than isolating radio's effect
- Not optimal for either city
"Eliminate newspaper advertising"
- What it says: Stop all newspaper advertising
- For testing the claim: This would test whether newspaper advertising affects sales, but doesn't directly test the consultant's claim about radio
- If sales remain the same, it could be due to radio or other factors - doesn't give a clear answer about radio specifically
"Eliminate radio advertising"
- What it says: Stop all radio advertising
- For testing the claim: This directly tests whether radio has any positive effect on sales
- Strong candidate for both cities - provides the most direct test of the consultant's claim
"Change the content of radio advertising"
- What it says: Keep radio advertising but modify the message
- For testing the claim: Introduces a new variable (content quality) that complicates the test
- Not optimal for testing the original claim
"Add television advertising"
- What it says: Begin TV advertising
- For testing the claim: Adds another confounding variable
- Makes the test less clear, not more clear
The Correct Answers
- For Part 1 (City X): Eliminate radio advertising
- For Part 2 (City Z): Eliminate radio advertising
By eliminating radio advertising in both cities, we create a situation where:
- Both cities have only newspaper advertising
- If sales remain equal → consultant is correct (radio doesn't improve sales)
- If sales decrease → consultant is wrong (radio does improve sales)
- This provides the most direct test of whether radio advertising has any positive effect
Common Traps to Highlight
Why not eliminate newspaper advertising?
- While this isolates the radio difference between cities, it doesn't directly test whether radio improves sales
- If sales remain the same, we can't definitively say whether it's because radio doesn't work or because radio alone is sufficient
- The consultant's claim is about whether radio improves sales, not about the differential between cities
Why not change strategies differently in each city?
- Making different changes in each city introduces new variables
- We want both cities to make the same change to maintain a controlled comparison
Why not add more advertising?
- Adding advertising (newspaper, TV) creates more confounding variables
- We need to simplify, not complicate, to isolate radio's effect
