See discussions, stats, and author profiles for this publication at: [Link]
net/publication/265551424
Quantitative information processing of Nutrition Facts panels
Article in British Food Journal · July 2014
DOI: 10.1108/BFJ-02-2013-0042
CITATIONS READS
19 776
1 author:
Lisa M Soederberg Miller
University of California, Davis
69 PUBLICATIONS 3,102 CITATIONS
SEE PROFILE
All content following this page was uploaded by Lisa M Soederberg Miller on 11 September 2014.
The user has requested enhancement of the downloaded file.
The current issue and full text archive of this journal is available at
[Link]/[Link]
Processing
Quantitative information of NFPs
processing of nutrition
facts panels
Lisa M. Soederberg Miller 1205
Department of Human Ecology, University of California, Davis,
Received 20 February 2013
California, USA Revised 2 June 2013
Accepted 4 June 2013
Abstract
Purpose – Quantitative information on nutrition labels (nutrition facts panels (NFPs)) is often conveyed
in the form of absolute weights (i.e. milligrams or grams) and reference values (i.e. per cent daily values
(%DVs)) which provide information regarding nutrient levels within the context of a total daily diet.
Some evidence suggests that %DVs are preferred over the weights and may be better at communicating
nutrition information. However, age differences are often neglected in past work, thus limiting the
understanding how effectively NFPs communicate quantitative information across adulthood. The paper
aims to discuss these issues.
Design/methodology/approach – Using eye tracking methodology, the present study examined
age and preference differences in attention to weights and %DVs on NFPs during two healthfulness
judgement tasks. The extent to which attention predicted judgement accuracy was explored and
findings were compared to two additional predictors, prior knowledge and NFP numeracy skills.
Findings – Although individuals paid attention to both types of quantitative information, attention to
%DVs, but not weights, predicted accuracy, on both tasks. For older adults only, preferences for %DVs
were related to %DV attention, and this in turn supported accuracy on the single-NFP task.
Originality/value – These data show that %DVs are important for healthfulness judgements across age
but that preferences for %DVs, together with attention to %DVs, are particularly important for older adults.
Keywords Nutrition, Older consumers, Nutrition labelling, Food labels, Quantitative information,
Prior knowledge
Paper type Research paper
Introduction
Diet plays a central role in maintaining health and reducing the risk of several chronic
diseases such as heart disease, some forms of cancer, and type 2 diabetes (Department
of Agriculture, US, 2004; US Department of Health and Human Services, 2000)
and nutrition facts panels (NFPs) on packaged foods contain important information
about the nutritional attributes (Campos et al., 2011; Cowburn and Stockley, 2004;
Grunert and Wills, 2007; Hieke and Taylor, 2012; Mhurchu and Gorton, 2007). However,
relatively little is known about how quantitative information in NFPs is processed.
Quantitative information is often presented in two metrics, absolute weights in
milligrams or grams and per cent daily values (%DVs), which place the nutrient values
within the context of total daily diet. However, few studies have examined how
individuals use quantitative information to make healthfulness decisions. This is an
important question because it underlies one of the most commonly reported uses of
NFPs, namely, to identify high and low nutrients (Guthrie et al., 1999). Moreover,
because of barriers associated with eating healthy diets in later life, including diet-related
British Food Journal
Vol. 116 No. 7, 2014
r Lisa Miller. pp. 1205-1219
Emerald Group Publishing Limited
This research was funded in part by a grant from the National Institutes of Health, R01AG19196. 0007-070X
The author wishes to thank the anonymous reviewers. DOI 10.1108/BFJ-02-2013-0042
BFJ chronic conditions such as hypertension and diabetes, social isolation, and dental problems
116,7 (Sahyoun et al., 2005), it is important to understand how quantitative information is
processed among older, as well as, younger adults. Consumer preferences for certain
types of information have been shown to be important influences on how food labels are
processed (Berning et al., 2010) but their influence specifically on older adults’ label
comprehension is unknown.
1206 The present study presents a model, outlined in Figure 1, linking preferences,
use of food label information, and comprehension to explore quantitative information
processing among younger and older adults. The framework is similar to past models
in which use of information and search intensity are important factors of nutrition-related
behaviour (Balasubramanian and Cole, 2002; Grunert and Wills, 2007). However, the
present model further specifies the role of attention as representing a form of information
use and the motivational role of preferences in directing attention to particular types
of information. That is, quantitative information is thought to be used effectively
if attention to this information is associated with accuracy of comprehension. The present
study tests the applicability of this model to younger and older adults and asks
whether age and preferences for quantitative information influences attention paid
to %DVs or weights on NFPs; and the effectiveness of this attention, that is, its ability to
predict accuracy of healthfulness judgement. Attention was assessed via eye tracking
methodology as consumers evaluated the healthfulness of foods represented only by
their NFPs. Eye tracking methodology has been used in several studies to examine
how consumers process information on food labels (Graham and Jeffery, 2011; van Herpen
and van Trijp, 2011; Visschers et al., 2010; Bialkova and van Trijp, 2011; Jones and
Richardson, 2007; Goldberg et al., 1999). Finally, nutrition knowledge and numeracy,
two abilities that are important for NFP comprehension (Grunert and Wills, 2007;
Rothman et al., 2006; Balasubramanian and Cole, 2002), were also examined to determine
if they could predict healthfulness accuracy over and above age, preferences, and attention
to quantitative information.
Preferences
Age for Weights or
%DVs
Attention to (Use of)
Weights, %DVs
Figure 1.
Model of proposed
relationships among
factors influencing Understanding
Nutrition Knowledge
quantitative information (Healthfulness
NFP Numeracy
processing on NFPs Evaluation Accuracy)
Preferences and use of quantitative information on NFPs Processing
Food labels on packaged foods often contain weight and reference values to provide of NFPs
consumers with an informative assessment of nutritional attributes of foods (Campos
et al., 2011). In the USA, NFPs are required (per the Nutrition Labelling and Education
Act) to present information on nutrients that have health benefits, such as calcium
(may reduce the risk of osteoporosis) and fibre (may reduce the risk of heart disease),
and nutrients to limit, such as trans fat, saturated fat, cholesterol, and sodium, which have 1207
been associated with chronic disease. Although quantitative information is an essential
aspect of communicating health information, research suggests that consumers have
difficulty interpreting quantitative facts on food labels and may overestimate their
abilities (Sharf et al., 2012; Campos et al., 2011; Cowburn and Stockley, 2004).
Absolute metrics (i.e. weight) may be particularly difficult to use relative to %DVs
(Levy and Fein, 1998) because different units are associated with different nutrients
(e.g. milligrams for sodium vs grams for fat) which could be compounded by lack of
numeracy skills, such as understanding how three grams compares to 300 milligrams.
Consumers may not know that the same amount of one nutrient may not have the same
healthful quality as the same amount of another (e.g. five grams of fibre vs five grams
of fat) (Levy and Fein, 1998; Taylor and Wilkening, 2008).
Per cent DVs, on the other hand, represent nutrient values based on the same scale,
one that considers the total daily diet. Because nutrient levels presented as %DVs are
self-anchoring in that individual values serve as a reference for the others (Taylor and
Wilkening, 2008), they could reduce the need to pay attention to weights when
assessing healthfulness. Some support for this comes from research manipulating
nutrition label formats and the type of quantitative information presented (Levy et al.,
1992, 1996). Levy and colleagues found that labels with %DVs helped consumers to
more readily identify high and low nutrient levels; however, it was the least preferred
format of the seven tested. Importantly, stated preferences for food label information
may not reflect what consumers are able to use effectively (Grunert et al., 2010).
Other evidence suggests that %DVs do not effectively communicate nutrition
information (Levy et al., 2000; Block and Peracchio, 2006). In a study of older adults at
risk of developing osteoporosis, researchers reported that patients had difficulty using
%DVs to determine the amount of calcium they needed in a day (Block and Peracchio,
2006). Even for a potentially easier task of judging healthfulness, %DVs are not always
used effectively (Barone et al., 1996; Li et al., 2000; Lubman et al., 2012; Borra, 2006).
Barone and colleagues (1996) found that the presence of %DVs had little effect on
healthfulness judgements. However, in that study, label formats were highly simplified,
making it unclear if %DVs were needed. Thus, %DVs may provide greater support
within the context of the more detailed NFPs that are present on food labels in the USA
and many other countries today. Moreover, the type of quantitative information used to
evaluate healthfulness may depend on the specific task, for example, when comparing
two food labels or evaluating one by itself (Barone et al., 1996; Viswanathan and
Hastak, 2002).
Age and food label comprehension
Past work suggests that age may impact processing of nutrition information on food
labels (Cole and Balasubramanian, 1993; Cole and Gaeth, 1990; Moorman, 1996; Burton
and Andrews, 1996; Miller and Cassady, 2012; Cowburn and Stockley, 2004). Still, as Hieke
and Taylor (2012) argue in their review of nutrition label comprehension, the evidence
is mixed, leaving it unclear whether and how age impacts NFP comprehension.
BFJ For example, Cole and Gaeth (1990) found age-related declines in the ability to apply
116,7 a set of rules to select the healthiest cereal from three alternatives. Importantly, the
reported means suggest that age differences decreased with training, suggesting that
experience and familiarity using NFPs may be an important factor in how effectively
quantitative nutrition information is used among older adults, as has been shown to be
the case among younger adults (Cowburn and Stockley, 2004; Moorman, 1996).
1208 Experience can also be considered within a larger time scale. Older individuals in
the USA have been exposed to the current form of NFPs for a longer time period than
have younger adults. The current design that includes %DV was specified by the US
Nutrition Labelling and Education Act in 1990, when younger adults were unlikely to
be making purchasing decisions.
Associations with nutrition knowledge and NFP numeracy
Knowledge has also been shown to be related to food label comprehension across
a range of adult ages (Grunert et al., 2010; Misra, 2007; Cowburn and Stockley, 2004;
Balasubramanian and Cole, 2002). Li et al. (2000) found that knowledge was associated
with preferences for %DVs as well as more effective use of %DVs. However, knowledge
in that study was manipulated prior to testing and may not represent knowledge that is
accumulated across time, which may be beneficial to older adults (Salthouse, 2003).
Past research has not examined the effect of knowledge as it relates to quantitative
information processing, making it unclear whether knowledge is related to accuracy and,
if so, whether attention, age, and preferences attenuate its effect.
Numeracy skills represent another potentially important influence on the quality of
quantitative information processing. Numeracy skills refer to “the ability to understand
and use numbers in daily life” (Rothman et al., 2006). There is some suggestion from the
health literacy literature that numeracy skills are central to understanding and using
nutrition and health information (Rothman et al., 2006). In the present study, numeracy
skills were operationalized specifically as they relate to the ability to read and manipulate
quantitative information on NFPs. As with knowledge, it is unclear whether this ability is
related to accuracy and how its role compares to attention, age, and preferences in its
ability to predict accuracy.
Present study
The present study explored the effects of age and preferences on first, the degree to
which individuals pay attention to weights and %DVs when evaluating healthfulness
of NFPs; and second, the effectiveness of this attention, assessed through examination
of associations between attention and accuracy. Other predictors of accuracy, namely,
prior knowledge and NFP numeracy, were included to determine if they could account
for variance over and above the key factors of age, preferences, and attention. Finally,
two healthfulness tasks were included to encompass a wider range of quantitative
information processes associated with healthfulness judgements. One task required the
evaluation of quantitative information on one NFP and the other task required a relative
judgement based on two NFPs.
Method
Participants
The study utilized a convenience sample of 33 younger adults (ages 19-34, M ¼ 23.2;
SD ¼ 4.0) and 32 older adults (ages 61-78, M ¼ 68.4; SD ¼ 4.7), some of whom
participated in an earlier study (Miller and Cassady, 2012). There were no age
differences in experience using NFPs, t o1. Experience was assessed using the item: Processing
“How frequently do you use Nutrition Facts panels when buying a food for the first of NFPs
time?” (1 ¼ never; 4 ¼ always).
Older adults had lower visual acuity scores as assessed by a Snellen test,
M ¼ 0.04, SD ¼ 0.07, than did younger adults, M ¼ 0.09, SD ¼ 0.03, t(63) ¼ 3.5,
p o0.01. Still, all participants had visual acuity LogMAR scores in the normal vision
range. Older adults reported taking more prescription medications, M ¼ 1.8, SD ¼ 1.7, 1209
than did younger adults, M ¼ 0.4, SD ¼ 0.8, t(63) ¼ 4.3, p o0.01, and were more likely
to report having a chronic condition based on a checklist of seven conditions (with 0-7
as possible scores), M ¼ 0.19, SD ¼ 0.4, relative to younger adults who reported having
none, M ¼ 0, SD ¼ 0, t(63) ¼ 2.7, p o0.01. Working memory capacity, assessed using
a composite measure of the loaded sentence span task (Stine and Hindman, 1994) and the
computation span (Salthouse and Babcock, 1991), showed that older adults, M ¼ 5.3,
SD ¼ 1.1, had a smaller working memory capacity than did younger adults, M ¼ 5.9,
SD ¼ 1.2, t(63) ¼ 2.2, p o0.05. However, older adults were high functioning as indicated
by years of education, M ¼ 17.8, SD ¼ 2.7, relative to younger adults, M ¼ 15.8, SD ¼ 1.6,
t(63) ¼ 3.6, p o0.01, and self-reported general health, M ¼ 1.3, SD ¼ 0.5, relative to
younger adults M ¼ 1.8, SD ¼ 0.7, t(63) ¼ 3.3, p o0.01. Health was rated on a scale of
1 (excellent) to 5 (poor); individuals in both age groups reported being in very good to
excellent health.
Age differences in key ability measures were also assessed. Older adults had higher
nutrition knowledge scores, M ¼ 0.56, SD ¼ 0.14, than did younger adults, M ¼ 0.46,
SD ¼ 0.22, t(63) ¼ 2.3, p o0.05. However, there were no age differences in numeracy, t o1.
NFPs
NFPs for a uniform, single-serving size were created using the NutritionData
web site ([Link]), which conforms to the US Food and Drug
Administration content, size, and format requirements that included black and white
fonts (21 CFR 101.9 (d)). The Appendix contains a sample NFP. NFPs from four
categories of foods (bread, soup, meat, and vegetables), commonly used in research
(Bingham et al., 1994; Schlettwein-Gsell et al., 1999) were included with the goal of
capturing a wide range of nutrient values. However, this categorical information was
not provided to the participant for either of the two tasks performed.
For the first task, participants viewed a single NFP and rated healthfulness on
a scale of 1 (very unhealthy) to 4 (very healthy). NFPs for bread and vegetable products
were lower in sodium and fat and were higher in fibre than were soup and meat NFPs.
Thus, scores of 3 or 4 (healthy) for bread and vegetable NFPs were scored as correct
and scores of 1 or 2 (unhealthy) were scored as correct for soup and meat NFPs. For the
second task, each NFP was paired with another that was significantly higher or lower
(by 75 per cent) on one nutrient but varied minimally and in the opposite direction
(5 per cent change) on all other nutrients. Participants were asked to select the NFP
that was healthier. The NFP containing the 75 per cent change in the healthy direction
was scored as the correct answer.
Attention was operationalized in terms of proportion dwell time in NFP interest
areas consisting of weights (gram or milligram), and %DVs corresponding to
cholesterol, fibre, saturated fat, total fat, sodium, and total carbohydrate, and weight
only for protein, sugar, and trans fats. With the exception of sugar and protein, these
core nutrients offered a choice between two types of quantitative information. Nutrient
names and information in the upper portion of the panel (i.e. amount calories and
BFJ calories from fat per serving, serving size information) or lower portion of the panel
116,7 (i.e. vitamins and minerals, total diet information) were not included as interest areas.
Due to age-related declines in speed of processing (Salthouse, 1996), age differences in
raw dwell time would not necessarily be driven by greater reliance on quantitative
information per se. Therefore, proportion dwell time was preferred over raw dwell
time. There were nine weight interest areas and six %DV interest areas (because
1210 protein, sugar, and trans fats are not accompanied by %DVs on NFPs in the USA).
Thus, proportion dwell times were divided by nine for weights and by six for %DVs.
The resulting measure represents the average proportion of time individuals spent
viewing the two types of quantitative information per trial, normed on the number of
possible interest areas for that information type.
NFP preferences for type of quantitative information
Preferences for quantitative information type were assessed using a question based on the
National Health and Nutrition Examination Survey food label use question (Centers for
Disease Control and Prevention, 2012), “What metric do you prefer when seeking
information in the nutrition facts panel?” Responses were coded by assigning those who
preferred %DVs into a %DV preference group and those who did not (i.e. indicated no
preference for one type of information over the other), into a weight preference group.
Prior nutrition knowledge
Knowledge of nutrition was assessed using a modified version of a multiple-choice,
38-item test of nutrition knowledge that was developed for an earlier study (Miller et al.,
2011). The measure contains semantic knowledge such as the relationships between
nutrition and health and knowledge of nutrition principles (e.g. which foods are good
sources of various nutrients), as well as procedural knowledge (e.g. knowing what to look
for when shopping for cereal). For this study, however, we excluded three items that
assessed numerical skills in using an NFP so they could be examined separately
(described next). The proportion of correct responses out of 35 was the score used to
reflect knowledge of nutrition.
NFP numeracy
The numeracy task consisted of three items requiring participants to view an NFP and
answer questions that required interpreting nutrient values and/or calculating
numerical information presented in the panel (e.g. determine the percentage of your
daily calcium needs are met if you ate two cups of the product shown). Scores on these
items were summed to produce an overall score.
Procedure
Participants completed a background demographic and health questionnaire followed by
the eye tracking tasks and individual-difference measures. Participants received a brief
training exercise in which the experimenter reviewed each of the key nutrient categories,
amounts (grams, milligrams), and %DV on a sample NFP printed on an 11 14 sheet
of paper. For the single-NFP task, participants indicated, on a scale of 1 (very unhealthy)
to 4 (very healthy), the healthfulness of one NFP presented on the screen by pressing the
corresponding button. For the two-NFP task, participants determined which of two NFPs,
shown simultaneously, was healthier. Participants pressed the left-most button to select
the left NFP or the right-most button to select the right NFP. Healthier NFPs were
counterbalanced across left and right positions.
Results Processing
Analytic approach of NFPs
There roughly equal numbers of younger (n ¼ 11) and older adults (n ¼ 10) in the %DV
preference group and equal numbers of younger (n ¼ 22) and older adults (n ¼ 22) in
the weights preference group, w2(65) ¼ 0.03, p 40.10. A mixed analysis of variance
was used to assess age, preference group, and task differences in attention to the two
types of quantitative information. Regression analyses predicting accuracy from 1211
attention, age, preferences, and ability were used to examine the effectiveness of
attention to quantitative information in the two tasks, relative to the other predictors.
Age and preference effects on attention
Mean proportion dwell time for quantitative information was analysed in a 2(Age: Young,
Old) 2(Preference for Information Type: Weight, %DV) 2(Attention to Information
Type: Weight, %DV) 2(Task: Single NPF, Double NPF) ANOVA with Attention to
Information Type and Task as the within-subjects variables. There was a significant
effect of Task, F(1, 61) ¼ 26.6 p o0.001, Z2 ¼ 0.30, showing that individuals paid more
attention per trial to quantitative information when comparing to two panels than
when evaluating one panel. A significant main effect of Attention to Information Type,
F(1, 61) ¼ 5.8, p o0.05, Z2 ¼ 0.09, showed greater attention to %DVs than to weights.
This effect was qualified by an Attention to Information Type Task interaction,
F(1, 61) ¼ 10.1 p o0.01, Z2 ¼ 0.14, indicating that more attention was paid to %DVs than
to weights when comparing two panels, t(64) ¼ 2.7, p o0.01, but not when evaluating
a single panel, t o1.
There were no main effects of Age, or Preference, F o1, for both, and the
Age Preference interaction was nonsignificant, F(1, 61) ¼ 2.7, p ¼ 0.10, Z2 ¼ 0.04.
However, there was a significant Age Preference Attention interaction,
F(1, 61) ¼ 4.7, p o0.05, Z2 ¼ 0.07. Figure 2 shows that older adults who preferred
%DVs allocated more attention to %DVs than to weights, t(9) ¼ 5.1, p o0.001,
however, older adults who preferred weights, and both groups of younger adults,
allocated attention equally to both types of quantitative information, t o1, for all.
Predictors of accuracy
Zero-order correlations were conducted to examine relationships between attention
and accuracy prior to controlling for other variables. As shown in Table I, attention
to %DVs, but not weights, was associated with accuracy on both tasks, suggesting
that attention to reference values is effective, but attention to weights is not.
Also, age and knowledge were positively associated with performance on the
double-NFP task, but not the single-NFP task. Given attention to weights failed to
predict accuracy on either task, no further models testing the effectiveness of
attention to weights were included.
Effectiveness of attention to %DVs was explored using hierarchical linear regressions
predicting accuracy from attention separately for each task. To address the question of
whether age and/or preferences interacted with attention to predict accuracy, a series of
cross-product terms were created between these three variables. Variables were entered in
three steps. In the first step, mean-centred main effects age and preferences were entered
along with the Age Preference term. In the second step, attention variables were added:
the main effects of attention and all interaction terms (Age Attention,
Age Preference, Preference Attention, Age Preference Attention). In the last
step, knowledge and numeracy were added.
BFJ Younger Adults
0.05
116,7 Attention to Weights
Attention to %DVs
Proportion Dwell TIme per Unit of Information
1212 0.03
0.00
Preference for Weights Preference for %DVs
Older Adults
0.05
Attention to Weights
Attention to %DVs
0.03
0.00
Preference for Weights Preference for %DVs
Figure 2.
Proportion dwell time for Note: Proportion dwell time is calculated as the average
%DVs and weights as a proportion of time spent on quantitative information per trial,
function of age and
preference group divided by the number of possible interest areas (six for %DV
and nine for weights)
Accuracy
Predictors Single-NFP task Double-NFP task
Age 0.15 0.29*
Preferences 0.13 0.16
Table I.
Knowledge 0.15 0.45**
Bivariate correlations
Numeracy 0.28* 0.32**
among age, preference
Attention to Weights 0.00 0.05
group, attention, prior
Attention to %DVs 0.35** 0.25*
knowledge, numeracy,
and accuracy Notes: *,**Correlation significant at 0.05 and 0.01 levels respectively (two-tailed)
As shown in Table II, the main effects of age and preferences and their cross-product term
were nonsignificant for the regression predicting accuracy on the single-NFP task.
In step 2, however, the main effect of attention was significant as were the Age Preference
and Age Preference Attention interactions. To examine the three-way interaction,
accuracy was regressed onto attention to %DVs to obtain the residuals representing
Single-NFP task Double-NFP task
Processing
b t b t of NFPs
Step 1
Age 0.15 1.22 0.25 2.07*
Preferences 0.15 1.22 0.16 1.25
Age Preferences 0.13 1.06 0.01 0.08 1213
Step 2
Attention to %DVs 0.34 2.79** 0.25 1.87***
Age Attention 0.20 1.61 0.13 0.76
Preferences Attention 0.04 0.34 0.11 0.68
Age Preferences Attention 0.25 2.02* 0.01 0.06 Table II.
Step 3 Regression models
Numeracy 0.20 1.64 0.131 1.06 predicting accuracy
Knowledge 0.13 1.08 0.382 3.13** from attention to
%DVs, age, preferences,
Notes: *,**,***Significant at po0.05; po0.01; po0.10 levels, respectively and ability by task
accuracy when attention is statistically removed. Follow up t-tests showed age declines in
accuracy for the %DV preference group, t(19) ¼ 2.6, p o0.05, but not in the weights
preference group, to1. This interaction, shown in Figure 3, suggests that, for older
adults who prefer %DVs, their performance was supported by their attention to
%DV. The main effect of attention remained significant even after the three-way term
was added, b ¼ 0.34, po0.001, suggesting that the positive association between attention
to %DVs and accuracy across individuals remained after partialling out the joint
effects of age, preferences, and attention.
For the double-NFP task, there was a main effect of age, with a positive beta weight
showing that older adults had higher scores on accuracy than did younger adults.
However, older adults’ performance advantage was no longer significant after controlling
for knowledge, b ¼ 0.13, p40.10, in the last step. Knowledge but not numeracy was
significant when added to the model. However, attention and interaction terms
with attention failed to predict accuracy on this task. In step 2, attention to %DVs
was attenuated relative to the significant zero-order association, suggesting that its
association with accuracy overlapped with age and preferences.
Discussion
NFPs offer important quantitative information surrounding the healthful aspects
of packaged foods. However, there are few if any studies that have assessed attention
to different types of quantitative information on NFPs. The present study was
undertaken to learn more about how younger and older adults varying in preferences
process quantitative information in NFPs to make decisions regarding healthfulness of
foods. The findings show that individuals pay attention to both %DVs and weights
when making healthfulness decisions. Results also show that attention to reference
values is connected to accuracy of healthfulness evaluations, but that attention to
weights is not. However, as discussed below, the findings also show that there are age
and preference group differences in how attention is allocated during the single- and
double-NFP tasks as well as differences in the effectiveness of attention. The data
presented are generally consistent with the model presented in Figure 1 indicating that
perception and age influence how consumers use nutrition information on food labels.
BFJ 1
Younger Adults
116,7
Older Adults
0.8
Accuracy Adjusted for
Attention to %DVs
0.6
1214
0.4
0.2
Figure 3.
Accuracy (proportion
correct) on the single-NFP 0
task by age, controlling for Weights %DVs
attention to %DVs
Preferences
Ageing
There are several findings that shed light on quantitative information processing in
adulthood. First, younger and older adults devoted similar amounts of attention to
quantitative information overall. This finding is inconsistent with the notion that older
adults use a satisficing heuristic, reducing the amount of information that is sought
(Cole and Balasubramanian, 1993). Moreover, older adults showed no deficits in their
ability to make accurate decisions regarding healthfulness and, on the comparison task,
showed higher performance than younger adults. Past research has not provided a clear
picture of whether and when there are age differences on NFP comprehension tasks.
For example, younger-old and older-old adults can locate information on an NFP equally
well (Byrd-Bredbenner and Kiefer, 2000). However, tasks that require calculations may be
more likely to show age declines (Burton and Andrews, 1996). In the present study,
evaluations of healthfulness required attention to the quantitative information but did
not require calculations, which are typically found to be challenging (Cowburn and
Stockley, 2004). It is important to note that when knowledge was statistically removed,
older adults no longer showed higher performance on the double-NFP task. Thus, older
adults’ knowledge appears to be important for healthfulness accuracy, a finding that is
consistent with past research using a variety of nutrition comprehension and memory
tasks (Miller et al., 2010, 2011, 2013; Miller and Cassady, 2012).
However, when preferences for the type of information are considered, some age
differences in attention emerged. Specifically, older adults who favour using %DVs on
NFPs, pay more attention to %DVs than to weights. Indeed this group of older adults
allocate more time to %DVs than do younger adults who also have preferences for
%DV. Thus, preferences have a greater impact on how attention is allocated among
older than younger adults.
It is unclear why a similar pattern was not found for older adults who prefer
weights and younger adults who prefer either metric. Findings could be attributable to
factors surrounding how preferences were assessed in the present study. For example,
it could be that some adults indicated information type preferences on NFP for uses
other than healthfulness evaluations. Individuals may use weights, for example,
to determine if their diet is meeting the dietary requirements or restrictions for their
specific chronic condition. However, past research using a variety of comprehension
tasks has shown that preferences for information type do not align with performance Processing
(Levy et al., 1992). Thus, an interesting question for future research is why do some of NFPs
adults prefer %DVs and use them effectively?
Effectiveness of attention to quantitative information
There were significant correlations between accuracy and attention to %DVs but not
to weights, suggesting that attention to reference values is more effective than is 1215
attention to weights. The fact that both tasks showed this association suggests some
similarities between the two types of healthfulness evaluations. For example, both
tasks require an assessment of the nutrients, those that should be limited and those
that may have beneficial effects on health, and their amounts. Reference values may be
useful in scanning for values within a single-NFP that are high or low and thereby
facilitate a determination of healthfulness. This would require only basic knowledge of
which nutrients should be high and which should be low in a healthy diet. Similarly,
for the comparison task, reference values are helpful. They do not require that
consumers know, for example, whether a two gram change across two foods translates
into a large difference for that nutrient because the change in per cent daily value will
reflect the magnitude of this change for that nutrient (Taylor and Wilkening, 2008).
A closer look at the effectiveness of attention to %DVs, however, yielded a more
complex picture for the single-NPF task. When attention to %DVs was statistically
removed, older adults who preferred %DVs performed more poorly than younger adults
who preferred %DVs. This suggests that older adults’ preferences are associated with the
effectiveness of attention in a way that is not evident for the younger adults. It could be
that older adults who preferred %DVs depended more on attention for this task because
knowledge was less relevant. As discussed below, knowledge requirements for the two
tasks differed.
Knowledge, numeracy, and accuracy
Findings from the present study are consistent with past work suggesting that
knowledge (Cowburn and Stockley, 2004; Grunert et al., 2010; Misra, 2007) and numeracy
(Rothman et al., 2006) are important for nutrition label comprehension. However, past
research has not compared the role of knowledge and numeracy. Results showed
numeracy was associated with greater performance on both tasks; however, the
association was no longer significant after controlling for age, preferences, and
attention. Prior nutrition knowledge was related to accuracy on the double-NFP task,
even after controlling for all other variables, but it was not related to performance on
the single-NFP task. It could be that when evaluating healthfulness in isolation,
a simple scan of high and low %DVs leads the consumer to the important nutrients
and only some basic knowledge of nutrition is needed to evaluate which nutrients are
desirable and which should be limited. When comparing two food labels, on the other
hand, consumers have twice as much information to consider. Prior knowledge may
help to organize and retain the individual comparisons so that an accurate evaluation
can be made.
Limitations and future directions
It is important to note that the sample size was small which may underestimate
relationships among attention, accuracy, and ability measures. Although the sample
showed the typical age declines in terms of chronic conditions, medications, and
working memory capacity, the older adults in this sample were high functioning in
BFJ terms of years of education, self-reported health, and numeracy skills. Thus, these
116,7 findings may not generalize to other segments of the population and age differences in
accuracy reported here should be considered a best-case scenario.
It is also important to point out that calculations were not required to accurately judge
healthfulness in either task. Future work requiring calculations to accurately
evaluate healthfulness is needed to provide a more stringent test of the role of age and
1216 attention to quantitative information as well as the role of numeracy in accuracy.
Research using NFPs from other countries and different types of reference information
(Viswanathan and Hastak, 2002) are also needed to establish generalizability of the
findings presented here.
Implications
Roughly one half of consumers report reading a product’s label when purchasing a food for
the first time (Federal Drug Administration, 2010), regardless of age (Ollberding et al.,
2010). One reason for low use could be challenges associated with processing and
understanding quantitative information (Levy and Fein, 1998; Levy et al., 1992, 1996;
Lewis and Yetley, 1992; Mhurchu and Gorton, 2007; Cowburn and Stockley, 2004). A recent
review of front-of-package labels concluded that information presented in the form of
percentages presents a challenge for consumers and should be avoided (Hawley et al.,
2013). However, evidence presented here suggests that weights may be less useful in
yielding a correct judgement than %DV information. Levy et al. (1992) also found that
label formats containing percentages lead to higher accuracy than other designs. Given the
benefits associated with using %DVs reported here, nutrition education efforts could be
targeted to increasing consumers’ preferences for, and effective use of, this type of
quantitative information on food labels.
References
Balasubramanian, S.K. and Cole, C. (2002), “Consumers’ search and use of nutrition information:
the challenge and promise of the nutrition labeling and education act”, The Journal of
Marketing, Vol. 66 No. 3, pp. 112-127.
Barone, M.J., Rose, R.L., Manning, K.C. and Miniard, P.W. (1996), “Another look at the impact of
reference information on consumer impressions of nutrition information”, Journal of Public
Policy & Marketing, Vol. 15 No. 1, pp. 55-62.
Berning, J.P., Chouinard, H.H., Manning, K.C., McCluskey, J.J. and Sprott, D.E. (2010), “Identifying
consumer preferences for nutrition information on grocery store shelf labels”, Food Policy,
Vol. 35 No. 5, pp. 429-436.
Bialkova, S. and van Trijp, H.C.M. (2011), “An efficient methodology for assessing attention to
and effect of nutrition information displayed front-of-pack”, Food Quality and Preference,
Vol. 22 No. 6, pp. 592-601.
Bingham, S.A., Gill, C., Welch, A., Day, K., Cassidy, A., Khaw, K.T., Sneyd, M.J., Key, T.J.A., Roe, L.
and Day, N.E. (1994), “Comparison of dietary assessment methods in nutritional
epidemiology: weighed records v 24 h recalls, food-frequency questionnaires and estimated-
diet records”, British Journal of Nutrition, Vol. 72 No. 4, pp. 619-643.
Block, L.G. and Peracchio, L.A. (2006), “The calcium quandary: how consumers use nutrition
labels”, Journal of Public Policy & Marketing, Vol. 25 No. 2, pp. 188-196.
Borra, S. (2006), “Consumer perspectives on food labels”, The American Journal of Clinical
Nutrition, Vol. 83 No. 5, p. 1235S.
Burton, S. and Andrews, J.C. (1996), “Age, product nutrition, and label format effects on consumer
perceptions and product evaluations”, Journal of Consumer Affairs, Vol. 30 No. 1, pp. 68-89.
Byrd-Bredbenner, C. and Kiefer, L. (2000), “The ability of elderly women to perform nutrition Processing
facts label tasks and judge nutrient content claims”, Journal of Nutrition for the Elderly,
Vol. 20 No. 2, pp. 29-46. of NFPs
Campos, S., Doxey, J. and Hammond, D. (2011), “Nutrition labels on pre-packaged foods:
a systematic review”, Public Health Nutrition, Vol. 14 No. 8, pp. 1496-1506.
Centers for Disease Control and Prevention (2012), “National health and nutrition examination
survey: 2007-2008 data documentation, codebook, and frequencies”, available at: [Link]. 1217
gov/nchs/nhanes/nhanes2007-2008/CBQPFC_E.htm
Cole, C.A. and Balasubramanian, S.K. (1993), “Age differences in consumers’ search for information:
public policy implications”, Journal of Consumer Research, Vol. 20 No. 1, pp. 157-169.
Cole, C.A. and Gaeth, G.J. (1990), “Cognitive and age-related differences in the ability to use
nutritional information in a complex environment”, Journal of Marketing Research, Vol. 27
No. 2, pp. 175-184.
Cowburn, G. and Stockley, L. (2004), “Consumer understanding and use of nutrition labelling:
a systematic review”, Public Health Nutrition, Vol. 8 No. 1, pp. 21-28.
Department of Agriculture, US (2004), “Quality of diets of older Americans Nutrition Insight 29”,
Center for Nutrition Policy and Promotion, Alexandria, VA.
Federal Drug Administration (2010), “Survey shows gains in food-label use, health/diet awareness”,
available at: [Link]/ForConsumers/ConsumerUpdates/[Link] (accessed May
2010).
Goldberg, J.H., Probart, C.K. and Zak, R.E. (1999), “Visual search of food nutrition labels”, Human
Factors: The Journal of the Human Factors and Ergonomics Society, Vol. 41 No. 3, pp. 425-437.
Graham, D.J. and Jeffery, R.W. (2011), “Location, location, location: eye-tracking evidence that
consumers preferentially view prominently positioned nutrition information”, Journal of
the American Dietetic Association, Vol. 111 No. 11, pp. 1704-1711.
Grunert, K. and Wills, J. (2007), “A review of European research on consumer response to
nutrition information on food labels”, Journal of Public Health, Vol. 15 No. 5, pp. 385-399.
Grunert, K.G., Wills, J.M. and Fernandez-Celemı́n, L. (2010), “Nutrition knowledge, and use and
understanding of nutrition information on food labels among consumers in the UK”,
Appetite, Vol. 55 No. 2, pp. 177-189.
Guthrie, J.F., Derby, B.M. and Levy, A.S. (1999), “What people know and do not know about
nutrition”, in Frazao, E. (Ed.), America’s Eating Habits: Changes and Consequences, USDA/
ERS, Washington, DC, pp. 243-280.
Hawley, K., Roberto, C.A., Bragg, M., Liu, P., Schwartz, M. and Brownell, K.D. (2013), “The science
on front-of-package food labels”, Public Health Nutrition, not in print form yet, doi: 10.1017/
S1368980012000754.
Hieke, S. and Taylor, C.R. (2012), “A critical review of the literature on nutritional labeling”,
Journal of Consumer Affairs, Vol. 46 No. 1, pp. 120-156.
Jones, G. and Richardson, M. (2007), “An objective examination of consumer perception of
nutrition information based on healthiness ratings and eye movements”, Public Health
Nutrition, Vol. 10 No. 3, pp. 238-244.
Levy, A.S. and Fein, S.B. (1998), “Consumers’ ability to perform tasks using nutrition labels”,
Journal of Nutrition Education and Behavior, Vol. 30 No. 4, pp. 210-217.
Levy, A.S., Fein, S.B. and Schucker, R.E. (1992), “More effective nutrition label formats are not
necessarily preferred”, Journal of the American Dietetic Association, Vol. 92 No. 10, pp. 1230-1234.
Levy, A.S., Fein, S.B., Schucker, R.E. (1996), “Performance characteristics of seven nutrition label
formats”, Journal of Public Policy & Marketing, Vol. 15 No. 1, pp. 1-15.
Levy, L., Patterson, R., Kristal, A. and Li, S. (2000), “How well do consumers understand
percentage daily value on food labels?”, Am J Health Promot, Vol. 14 No. 3, pp. 157-160.
BFJ Lewis, C.J. and Yetley, E.A. (1992), “Focus group sessions on formats of nutrition labels”, Journal
of the American Dietetic Association, Vol. 92 No. 1, pp. 62-66.
116,7
Li, F., Miniard, P.W. and Barone, M.J. (2000), “The facilitating influence of consumer knowledge
on the effectiveness of daily value reference information”, Journal of the Academy of
Marketing Science, Vol. 28 No. 3, pp. 425-435.
Lubman, N., Doak, C. and Jasti, S. (2012), “Food label use and food label skills among immigrants
1218 from the Former Soviet Union”, Journal of Nutrition Education and Behavior, Vol. 44 No. 5,
pp. 398-406.
Mhurchu, C.N. and Gorton, D. (2007), “Nutrition labels and claims in New Zealand and Australia:
a review of use and understanding”, Australian and New Zealand Journal of Public Health,
Vol. 31 No. 2, pp. 105-112.
Miller, L.M.S. and Cassady, D.L. (2012), “Making healthy food choices using nutrition facts
panels: the roles of knowledge, motivation, dietary modifications goals, and age”, Appetite,
Vol. 59 No. 1, pp. 129-139.
Miller, L.M.S., Gibson, T.N. and Applegate, E.A. (2010), “Predictors of nutrition information
comprehension in adulthood”, Patient Education and Counseling, Vol. 80 No. 1, pp. 107-112.
Miller, L.M.S., Zirnstein, M. and Chan, P. (2013), “Memory for nutrition information: the roles of
age and prior knowledge in recognition”, Journal of Health Psychology.
Miller, L.M.S., Gibson, T.N., Applegate, E.A. and de Dios, J. (2011), “Mechanisms underlying
comprehension of health information in adulthood: the roles of prior knowledge and
working memory capacity”, Journal of Health Psychology, Vol. 16 No. 5, pp. 794-806.
Misra, R. (2007), “Knowledge, attitudes, and label use among college students”, Journal of the
American Dietetic Association, Vol. 107 No. 12, pp. 2130-2134.
Moorman, C. (1996), “A quasi experiment to assess the consumer and informational
determinants of nutrition information processing activities: the case of the nutrition
labeling and education act”, Journal of Public Policy & Marketing, Vol. 15 No. 1, pp. 28-44.
Ollberding, N.J., Wolf, R.L. and Contento, I. (2010), “Food label use and its relation to dietary intake
among US adults”, Journal of the American Dietetic Association, Vol. 110 No. 8, pp. 1233-1237.
Rothman, R.L., Housam, R., Weiss, H., Davis, D., Gregory, R., Gebretsadik, T., Shintani, A. and
Elasy, T.A. (2006), “Patient understanding of food labels: the role of literacy and numeracy”,
American Journal of Preventive Medicine, Vol. 31 No. 5, pp. 391-398.
Sahyoun, N.R., Zhang, X.L. and Serdula, M.K. (2005), “Barriers to the consumption of fruits and
vegetables among older adults”, Journal of Nutrition for the Elderly, Vol. 24 No. 4, pp. 5-21.
Salthouse, T. (2003), “Interrelations of aging, knowledge, and cognitive performance”, in
Staudinger, U.M. and Lindenberger, U. (Eds), Understanding Human Development:
Dialogues with Lifespan Psychology, Kluwer, Dordrecht, pp. 265-287.
Salthouse, T.A. (1996), “The processing speed theory of adult age differences in cognition”,
Psychological Review, Vol. 103 No. 3, pp. 403-428.
Salthouse, T.A. and Babcock, R.L. (1991), “Decomposing adult age differences in working
memory”, Developmental Psychology, Vol. 27 No. 5, pp. 763-776.
Schlettwein-Gsell, D., Decarli, B. and De Groot, L. (1999), “Meal patterns in the SENECA study of
nutrition and the elderly in Europe: assessment method and preliminary results on the role
of the midday meal”, Appetite, Vol. 32 No. 1, pp. 15-22.
Sharf, M., Sela, R., Zentner, G., Shoob, H., Shai, I. and Stein-Zamir, C. (2012), “Figuring out food
labels. Young adults’ understanding of nutritional information presented on food labels is
inadequate”, Appetite, Vol. 58 No. 2, pp. 531-534.
Stine, E.A.L. and Hindman, J. (1994), “Age differences in reading time allocation for
propositionally dense sentences”, Aging and Cognition, Vol. 1 No. 1, pp. 1-12.
Taylor, C.L. and Wilkening, V.L. (2008), “How the nutrition food label was developed, part 1: the Processing
nutrition facts panel”, Journal of the American Dietetic Association, Vol. 108 No. 3, pp. 437-442.
of NFPs
US Department of Health and Human Services (2000), Healthy People 2010: Understanding and
Improving Health, US Department of Health and Human Services, Washington, DC.
van Herpen, E. and van Trijp, H.C.M. (2011), “Front-of-pack nutrition labels. Their effect on
attention and choices when consumers have varying goals and time constraints”, Appetite,
Vol. 57 No. 1, pp. 148-160. 1219
Visschers, V.H., Hess, R. and Siegrist, M. (2010), “Health motivation and product design
determine consumers’ visual attention to nutrition information on food products”,
Public Health Nutrition, Vol. 13 No. 7, pp. 1099-1106.
Viswanathan, M. and Hastak, M. (2002), “The role of summary information in facilitating
consumers’ comprehension of nutrition information”, Journal of Public Policy & Marketing,
Vol. 21 No. 2, pp. 305-318.
About the author
Lisa M. Soederberg Miller is a Associate Professor of Human Development in the Department of
Human Ecology. Her research focuses on age differences in nutrition information processing in
adulthood. Pofessor Lisa M. Soederberg Miller can be contacted at: lmsmiller@[Link]
Appendix. Sample nutrition facts panel
Nutrition Facts
Serving Size 100g
Servings Per Container 1
Amount Per Serving
Calories 269 Calories from Fat 38
% Daily Value*
Total Fat 4g 7%
Saturated Fat 1g 4%
Trans Fat 0g
Cholesterol 0mg 0%
Sodium 599mg 25%
Total Carbohydrate 48g 16%
Dietary Fiber 4g 16%
Sugars 8g
Protein 8g
Vitamin A 0% • Vitamin C 0%
Calcium 7% • Iron 15%
*Percent Daily Values are based on a 2,000 calorie
diet. Your Daily Values may be higher or lower
depending on your calories needs:
Calories 2,000 2,500
Total Fat Less than 65g 80g
Sat Fat Less than 20g 25g
Cholesterol Less than 300mg 300mg
Sodium Less than 2400mg 2400mg
Total Carbohydrate 300g 375g
Dietary Fiber 25g 30g
To purchase reprints of this article please e-mail: reprints@[Link]
Or visit our web site for further details: [Link]/reprints
View publication stats