[PEREZGONZALEZ Jose D (2012). Marketing claims and the nutritional balance of breakfast cereals. Journal of Knowledge Advancement & Integration (ISSN 1177-4576), 2012, pages 97-104.] [Printer friendly]
Marketing claims and the nutritional balance of breakfast cereals
Perezgonzalez assessed the nutritional balance of breakfast cereals in 2012 (see b3,c4,d5,e6,f7,g8). The same data can be analysed further, to ascertain whether marketing claims may (unintentionally) inform about the overall nutritional balance of above products. The claims of greatest interest are nutrition-related claims (such as low-sodium, low-sugar, high-fiber, and low-fat), as well as more generic claims related to health or wellbeing (such as wholegrain, heart-health -eg, the Heart Foundation's "Tick"-, and overall health -eg, Pams' "Pick Me sun" logo).
Illustration 1 collates information per group within claims, namely group size, as well as nutritional balance (BNI) median and interquartile range. Illustration 2 collates information about mean ranks per group, as well as tests for equality of independent ranked distributions (Mann-Whitney U tests and equivalent Z areas).
|Illustration 1: Group size, median & interquartile range per claim group|
|n||median||( IQR )||n||median||( IQR )|
|High fiber||53||56.21||( 55.13 )||26||36.81||( 33.09 )|
|Low sodium||58||40.89||( 55.12 )||21||33.14||( 23.98 )|
|Low sugar||75||39.68||( 51.82 )||4||22.04||( 27.60 )|
|Low fat||39||35.40||( 50.28 )||40||44.65||( 43.44 )|
|Wholegrains||65||55.58||( 54.70 )||14||34.27||( 19.73 )|
|Pick Me sun||76||39.59||( 52.17 )||3||29.53||( … )|
|Heart health||56||36.99||( 61.22 )||23||40.89||( 27.27 )|
|[Tick||57||37.23||( 60.85 )||22||40.20||( 27.55)]|
|(Medians closer to 0.0 indicate greater balance.)|
According to these results, it can be concluded that:
- Cereals grouped according to 'wholegrain' claims show ranked group distances so extreme that, by chance alone, such distances would only occur about 4 times in a 1000. As 'wholegrain' cereals have a lower median (34.27) than otherwise (55.58), it can be concluded that cereals with a 'wholegrain' claim are more balanced nutritionally. Indeed, such claim alone would identify some 32% of more nutritionally balanced breakfast cereals (see rho values in illustration 3).
- Cereals grouped according to 'low sugar' claims show ranked group distances so extreme that, by chance alone, they would only occur about 5 times in a 100. As these cereals have a lower median (22.04) than otherwise (39.68), it can be concluded that cereals with a 'low sugar' claim are more balanced nutritionally. Indeed, such claim alone would identify some 22% of more nutritionally balanced breakfast cereals.
- Cereals grouped according to 'low fat' claims show ranked group distances so extreme that, by chance alone, they would only occur about 2 times in a 100. As these cereals have a higher median (44.65) than otherwise (35.40), it can be concluded that cereals with a 'low fat' claim are less balanced nutritionally. Indeed, such claim alone would identify some 26% of less nutritionally balanced breakfast cereals.
- Cereals with 'high fiber' and 'low sodium' claims, and, perhaps, those with the 'Pick Me sun' logo, may help identify the nutritional balance of breakfast cereals. However, this procedure is increasingly less dependable, for, by chance alone, group differences in nutritional balance would occur about 10 times in a 100 for cereals with 'high fiber' claims, about 13 times in a 100 for cereals with 'low sodium' claims, and about 19 times in a 100 for cereals with the 'Pick Me sun' logo. Furthermore, at most the nutritional balance of 19%, 17% and 15% of cereals would be identified with any of those claims, respectively.
- 'Heart health' claims (including the Heart Foundation's 'Tick' logo) inform practically nothing about the nutritional balance of breakfast cereals.
|Illustration 2. Mann-Whitney U tests|
|Claims||m.rank 'no'||m.rank 'yes'||U||Z||( p )|
|Wholegrain||45.15||29.50||416.0||-2.85||( .004 )|
|Low fat||36.48||49.71||405.0||-2.26||( .024 )|
|Low sugar||41.18||17.88||61.5||-1.98||( .048 )|
|High fiber||44.40||35.71||608.5||-1.68||( .093 )|
|Low sodium||41.81||31.61||337.5||-1.51||( .131 )|
|Pick Me sun||40.68||22.83||62.5||-1.32||( .186 )|
|Heart health||39.38||41.52||609.0||-.38||( .706 )|
|[Tick||39.68||40.82||609.0||-.20||( .844 )]|
|(Mean ranks closer to 0.0 indicate greater balance. 2-tailed significance tests)|
Illustration 3 collates information about three multiple regression models which explore combinations of claims. Overall, when considering all seven claims at once (regression model 1, R1, 'enter' procedure), the nutritional balance of 52% of breakfast cereals in the sample can be identified correctly (ie, more claims predict a more balanced product even if not all claims appear on the same product). The second regression model (R2, 'enter' procedure) eliminates the less relevant claims from model 1, and is able to identify the nutritional balance of about 52% of breakfast cereals in the sample by attending to just five claims: 'wholegrain', '[absence of] low fat', 'low sugar', the 'Pick Me sun' logo, and '[absence of] heart health-related claims'. The third regression model (R3, 'stepwise' procedure) retains the three claims found significant earlier (those presented in illustration 2), and is able to identify the nutritional balance of about 48% of breakfast cereals in the sample by attending to just three claims: 'wholegrain', 'low sugar', and '[absence of] low fat'. Each model is able to predict the nutritional balance of breakfast cereals in approximately the same proportion, but the latter models allow to do so by attending to a lesser number of claims, thus being more efficient. Indeed, the 'usability' of each regression model for predicting the nutritional balance of breakfast cereals not in the sample is, approximately, 45%, 47% and 45%, respectively (adjusted Rs).
|Illustration 3. Multiple regression models on the ranked BNI|
|rho||( p )||β||( p )||β||( p )||β||( p )|
|Wholegrain||-.323||( .004 )||-.356||( .003 )||-.380||( .001 )||-.318||( .002 )|
|Low fat||.256||( .023 )||.284||( .008 )||.271||( .010 )||.272||( .009 )|
|Low sugar||-.224||( .047 )||-.170||( .182 )||-.195||( .112 )||-.270||( .010 )|
|High fiber||-.190||( .093 )||-.034||( .789 )|
|Low sodium||-.171||( .132 )||-.069||( .569 )|
|Pick Me sun||-.150||( .188 )||-.156||( .222 )||-.169||( .180 )|
|Heart health||.043||( .708 )||.154||( .165 )||.159||( .141 )|
|R1 =||.524||( .001 )|
|R2 =||.517||( .000 )|
|R3 =||.482||( .000 )|
|Adj.R1 =||.451||Adj.R2 =||.467||Adj.R3 =||.449|
|(Dependent variable = BNI. 2-tailed significance tests)|
Jose D PEREZGONZALEZ (2012). Massey University, Turitea Campus, Private Bag 11-222, Palmerston North 4442, New Zealand. (JDPerezgonzalez).
Want to know more?
- BNI™ database
- The database offers individual nutrition analyses for foods, including the food referred to in above article.
- BNI™ journal (2012, issue 3) - Breakfast cereals
- This issue of the Balanced Nutrition Index™ journal collates all BNI™ nutrition information for the original sample in a single book.
- Wiki of Science - Nutritional balance of foods
- This Wiki of Science page collates information about several foods on a single page and provides useful links to the appropiate files.