Content
Citation(s) & Footnote(s)
Book(s)Bibliography and coverage
Albers, Alison Burke; DeJong, William; Naimi, Timothy S.; Siegel, Michael; and Jernigan, David H. (2014). The relationship between alcohol price and brand choice among underage drinkers: Are the most popular alcoholic brands consumed by youth the cheapest? Substance Use & Misuse 49(13), 1833–1843. DOI: 10.3109/10826084.2014.935790.
Fully read the supplied 21-page NIH author manuscript: narrative, methods, discussion, glossary, disclosures, funding, all 32 bibliography entries and all table rows/notes. Visually inspected all four numbered tables on PDF pages 14–21; Table 4 spans pages 17–21. No figures or separate appendix present. Page locators below refer to this supplied PDF, not final journal pagination. Publisher metadata checked; final publisher tables were not accessible for collation. Full supplied-source review does not resolve its editorial defects.
Research question and design — pages 1–7
The paper tests whether youth-popular brands are simply the cheapest. Survey fieldwork December 2011–May 2012: 1,032 initially sampled past-30-day drinkers aged 13–20, 1,031 in analysis. It reuses the respondent dataset of Siegel et al. (2013), so these are related analyses, not independent replications. Conditional response rates were about 43–44%; Black and lower-income invitees were less likely to respond. These are recent drinkers, not all youth and not adult whiskey customers.
2011 posted brand prices were compiled from control-state listings and online retailers. Researchers selected each outlet's least-expensive package/SKU for a brand, standardized to an ounce of pure ethanol, then averaged available prices. This is not the respondent's transaction price, a market-share-weighted price, or an estimate of acquisition costs when alcohol is shared/gifted. Taxes and access conditions limit comparability. The heading referring to standard drinks must not override the actual unit: dollars per ounce of pure ethanol. A US standard drink is 0.6 fluid ounces of ethanol; the two units are not interchangeable.
The price list includes 951 brands (898 listed plus 53 write-ins). The logistic outcome is membership in 27 brands with at least 0.87% consumption share. The top-25 prevalence ranking is a different grouping. Prevalence, drink share and popularity thresholds must remain distinct.
Findings with limits — pages 7–10
The reported overall odds ratio is 0.59 (95% CI 0.42–0.85), and popular brands have a reported mean price of $1.80 versus $3.40 for other brands. These are observational brand-level associations; they are not individual price elasticity or causal marketing effects. Reported beer and spirits mean differences were not statistically significant. Popular brands are frequently not the cheapest within their category, which is compatible with price being only one influence. The study does not directly measure branding exposure, access, motivations or brand availability, and cannot establish which explains the residual pattern. Nor does it overturn evidence that pricing policies can affect total consumption.
Do not use these historical youth data for current prices, adult demand forecasting, youth promotion, health reassurance or a claim that branding beats price.
Table and citation audit — pages 7–21
- Page 7 refers to Table 1 for mean-price comparisons, but supplied Table 1 (p.14) contains demographics. The referenced comparison table is not present as described. All 21 supplied pages are intact; this is a version/collation gap, not an inferred missing physical page.
- Table 2 (p.15) reports wine OR 0.22 with CI 0.66–0.93: the point estimate lies outside its interval. Do not silently repair it. Displayed category counts total 946 rather than 951; omission of five fortified wines is a possible explanation, not verified.
- Table 3 (p.16) labels its last column Consumption Prevalence Rank, while its note defines price rank. Its displayed 25 prices give mean 1.9128 (consistent with 1.91) but median 1.77, not the printed 1.74. This arithmetic is reviewer recalculation of displayed rounded values.
- Four Loko is $0.89 in the narrative but $0.82 in Tables 3–4; 89 is its displayed price rank. Smirnoff Malt Beverages ranks 28 within category in the narrative and 38 in Table 4.
- The narrative calls Natural Light the least-expensive top-ten beer, although Table 4 shows Keystone Light $1.18 and Natural Light $1.19.
- Corona Extra's overall rank appears as 3.83 in Table 4 versus 383 in Table 3.
- Jack Daniel's changes from $1.77/11.4% in Table 3 to $1.74/11.5% in Table 4; Bacardi changes $1.00 to $0.99, Hennessy 5.6% to 5.7%, and Miller 7.4% to 7.5%. Retain conflicts rather than choosing a convenient number.
- Table 4 fortified-wine prices and ranks do not order consistently: Night Train $1.04 is rank 1 while Cisco $0.91 is rank 2. The displayed values do not support a reliable corrected ranking.
- Narrative significance claims/category lists do not align cleanly with Table 4 asterisks (including brandy and whiskey). Do not infer corrected significance without data.
- The glossary (p.11) defines unpopular with a greater-than sign, contradicting the methods' less-than threshold. Pilot-study 2011 in-text dates differ from 2012 bibliography dates. A marketing-related Anderson citation points to the other lettered entry. These are citation/editing warnings, not evidence against every substantive conclusion.
The repeated discrepancies warrant low confidence in direct numeric reuse until the final edition or underlying data is reconciled.
Academy contribution and linked synthesis
Use internally as a worked example of auditing units, denominators, table arithmetic and observational inference. A proposed research-literacy exercise can ask adult learners why a low price rank does not demonstrate purchase motivation, and why a conditional youth survey cannot establish adult market demand. Do not reproduce flawed numbers in public lessons.
Pair with No access and the existing idea 
This critical analysis is complete for the supplied version. Final-edition table collation, Proton byte identity and browser rendering remain separately unverified. No public course changes made.
