Impact of hedonic evaluation on consumers’
preferences for beef enriched with Omega 3:
A Generalized Multinomial Logit Model approach
ID 371
Yasmina BABA1, Zein KALLAS2, José Maria GIL2, Carolina REALINI3
1INSTITUTO DE
SOSTENIBILIDAD, UPC, BARCELONA, SPAIN.
2CREDA, CASTELLDEFELS, SPAIN.
3IRTA, MONELLS, SPAIN.
MOTIVATION
 CONSUMERS usually face a COMPLEX DECISION process.
 MANY CHARACTERISTICS and several products' ALTERNATIVES.
 Such COMPLEXITY entangles the determination of the FACTORS that
AFFECT the consumer’s DECISION MAKING.
 HEALTH CONCERNS are one of the MOST IMPORTANT PREDICTORS for
food consumption.
 FAT CONTENT is receiving HIGHER ATTENTION over the last decades.
Zein.Kallas@upc.edu
MOTIVATION
 Food & beverage DEMANDS has EVOLVED towards new range of
products
often
related
to
HEALTH-PROMOTION
&
DISEASE
PREVENTION.
 Consumers increased their INTEREST in the HEALTH BENEFITS of
such products.
 The number of REDUCED-FAT FOODS is INCREASING.
 Foods with increased levels of UNSATURATED FATTY ACIDS
(OMEGA 3) are gaining MARKET SHARES (Milk, eggs…).
Zein.Kallas@upc.edu
FIRST OBJECTIVE
 SPANISH Consumers’ PREFERENCES toward BEEF MEAT ENRICHED with
POLYUNSATURATED FATTY ACIDS:
 Omega 3.
 Conjugated Linoleic Acid (CLA).
Zein.Kallas@upc.edu
FIRST OBJECTIVE
3. EMPIRICAL APPLICATION: Experimental design
CLA
Omega 3
3. EMPIRICAL APPLICATION: Experimental design
http://www.topsmarkets.com/departments/GreatO_Meats
MOTIVATION
 SENSORY ATTRIBUTES are also decisive factors for ACCEPTANCE of
food.
 TASTE (Flavour, odour …) are a primary basis to make REPEAT
PURCHASES of a product.
 HEALTH
and
NUTRITION
information
together
with
EATING
EXPERIENCE are all relevant for a POSITIVE valuation of specific
FUNCTIONAL FOOD.
Zein.Kallas@upc.edu
MOTIVATION
 Many studies have analysed consumers’ PREFERENCES, ATTITUDES
and ACCEPTANCE towards BEEF.
 However, LITERATURE that analyses HOW the HEDONIC VALUATION
(EATING AND SENSORY EXPERIENCE) affect consumers’ purchasing
decisions is STILL SCARCE and remains UNTREATED for BEEF
ENRICHED with POLYUNSATURATED FATTY ACIDS, particularly in
SPAIN.
Zein.Kallas@upc.edu
SECOND OBJECTIVE
 To analyse the IMPACT OF HEDONIC EVALUATION, for both
INFORMED and NON-INFORMED CONSUMERS, on the EXPECTED
PREFERENCES for BEEF ENRICHED with omega-3 and CLA by
comparing PREFERENCES BEFORE ante AFTER TASTING experience.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH
 We attempt to MIMIC CONSUMERS’ behaviour towards a novel
product which can be summarized in 3 MAIN SUBSEQUENT steps:
1. When consumers face an NEW PRODUCT on store shelves they
generate
EXPECTATIONS
EXPERIENCES
and
on
AVAILABLE
the
basis
of
INFORMATION
their

PAST
CHOICE
EXPERIMENTS (CE).
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH
2. Consumers TASTE the new product allows CONSTRUCTING A
CURRENT SENSORY EXPERIENCE that is useful to decide for a
repeated choice or not SENSORY TEST.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH
3. After tasting the new product, consumers’ acceptance may
result in AGREEMENT OR DISAGREEMENT with what they
EXPECTED. These changes play an important role in the final
choice of the consumers WE REPEAT THE SAME CHOICE
EXPERIMENT.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: First and the third steps, the CE
 Within the range of techniques that analyze consumers’ preferences,
we used The CHOICE EXPERIMENT (CE) due to its capacity to analyze
preferences for ‘COMPLEX GOODS’.
 The conceptual foundations of CE rely on two main theories:
a) LANCASTER’S THEORY OF VALUE which proposes that utilities for
goods can be DECOMPOSED into separable utilities for their
characteristics or ATTRIBUTES.
b) RANDOM UTILITY THEORY
which explains the DOMINANCE
JUDGMENTS made between pairs of offerings.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: First and the third steps, the CE
 The product is DESCRIBED through an array of ATTRIBUTES.
 The COMBINATION of attributes’ levels following orthogonal designs
allow to create hypothetical scenarios: “ALTERNATIVES”.
 Alternatives are ARRANGED together following experiment designs to
constitute “CHOICE SETS”.
 Respondent are ASKED to chose between the ALTERNATIVES in each
CHOICE SET.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: First and the third steps, the CE
Utility Theory
𝑈𝑗𝑛 = 𝑉𝑗𝑛 𝑋𝑗𝑛 , 𝑆𝑛 + 𝜀𝑛
Multinomial Logit Model
𝑈𝑛𝑗𝑡 = 𝛽𝑥𝑛𝑗𝑡 +
𝜀𝑛𝑗𝑡
𝜎𝑛 ; 𝑛 = 1. . 𝑁 𝑗 = 1. . 𝐽 𝑡 = 1. . 𝑇
 It imposes HOMOGENEITY in preferences for observed attribute.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: First and the third steps, the CE
Mixed Logit Model
𝑈𝑛𝑗𝑡 = 𝛽𝑛 𝑥𝑛𝑗𝑡 +
𝜀𝑛𝑗𝑡
𝜎𝑛 ; 𝑛 = 1. . 𝑁 𝑗 = 1. . 𝐽 𝑡 = 1. . 𝑇
 It accommodates for the UNOBSERVED HETEROGENEITY by
allowing RANDOM COEFFICIENTS on attributes.
 Recent STUDIES ARGUED that much of the PREFERENCE
HETEROGENEITY can be better captured by the SCALE.
 The MIXL turns to be likely a POOR APPROXIMATION if scale
heterogeneity is NOT ACCOUNTED FOR .
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: First and the third steps, the CE
Generalized Multinomial Logit Model
𝑈𝑛𝑗𝑡 = [𝜎𝑛 𝛽 +𝛾𝜂𝑛 +[ 1 − 𝛾 𝜎𝑛 𝜂𝑛 𝑋𝑛𝑗𝑡 + 𝜀𝑛𝑗𝑡
𝜸: A mixing parameter
between 0 & 1
𝑼𝒏𝒋𝒕 : The Utility to person 𝒏 from
choosing alternative j on choice set t
 The relative importance of
attributes are calculated
Error term
𝝈𝒏 : A scaling factor that
proportionally scales the 𝜷 for
each individual
Ik 
 max  k  min  k 
K
  max 
k 1
k
 min  k 
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: Empirical application of the CE
Attributes
 ORIGIN: locally produced or not.
 MEAT COLOUR: pale red or bright red meat.
 FAT CONTENT: moderate visible fat or slight visible fat.
 ENRICHMENT: conventional, enriched with Omega-3, with CLA and
with both Omega3 + CLA.
 PRICE: was defined in € per tray of 0.3 kg: 6.6€ as high, 5.7€ as
medium-high, 4.8€ as medium-low and 3.9€ as low.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: Empirical application of the CE
Experimental Design
 The full factorial design  a total of 128 (42x23) ‘alternatives’.
 3 alternative by choice sets (42x23)3 possible combinations .
 Follow a MAIN EFFECT ORTHOGONAL FRACTIONAL FACTORIAL
DESIGN: 16 choice sets.
 However, there were still too many questions for a single respondent
which is COGNITIVE burden and TIME consuming.
 We used FACTORIAL BLOCKING arrangement, obtaining two blocks,
each with 8 CHOICE SETS.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: Empirical application of the CE
Choice set
B
A
10 20683/A
CEE
Plot/Reference:
2611
Store between 0C y 4C
BEEF
STEAK
Born in:
Fattened in:
Slaughtered in:
Cut up in:
Spain
Spain
Spain
Spain
Price Kg Net weight
22.00 €
0.300 Kg.
Total €: 6.60 €
10 20683/A
CEE
Plot/Reference:
2611
Store between 0C & 4C
BEEF
STEAK
Born in:
Fattened in:
Slaughtered in:
Cut up in:
Catalonia
Catalonia
Catalonia
Catalonia
C
BEEF
STEAK
10 20683/A
CEE
Plot/Reference:
2611
Store between 0C & 4C
Born in:
Fattened in:
Slaughtered in:
Cut up in:
Price Kg Net Weight
13.00 €
Price Kg Net weight
16.00 €
0.300 Kg.
Total €: 3.90 €
1. Considering that “A”, “B” and “C” are the only available products, which product would you choose? “A” 
2. Would you purchase your chosen product? Yes 
No 
Catalonia
Catalonia
Catalonia
Catalonia
0.300 Kg.
Total €: 4.80 €
“B” 
“C” 
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: Empirical application of the CE
Data collection
 Face-to-face questionnaire completed in a
controlled
environment.
represents
consumers
The
who
population
REGULARLY
PURCHASE FOOD and beverages and having
PURCHASED BEEF MEAT IN THE LAST
MONTH.
PAMPLONA
(Basque Country)
240
BARCELONA
(Catalonia)
240
ZARAGOZA
(Rest of Spain)
240
 Quota sampling stratified by GENDER and
AGE.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: Empirical application of the CE
Data collection
Sample by CITY
240
Sample A
Sample B
120
120
( CE without information)
( CE with information)
Bloc 1
(60)
Bloc 2
(60)
Bloc 1
(60)
3 cities
Bloc 2
(60)
Sensory test
Informed about what they taste
Repeat the CE
Repeat the CE
120
120
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: The second step the sensory test
 They
taste
(BLIND
CONDITIONS)
4
TYPES
OF
BEEF:
1)
CONVENTIONAL, 2) ENRICHED with OMEGA3, 3) ENRICHED with CLA,
4) ENRICHED with both CLA and OMEGA3.
 Each consumer rated OVERALL LIKING using a 9-POINT SCALE (1
‘dislike extremely’, 2 ‘dislike very much’, 3 ‘dislike moderately’, 4
‘dislike slightly’, 5 ‘neither like nor dislike’ 6 ‘like slightly’, 7 ‘like
moderately’, 8 ‘like very much’, 9 ‘like extremely’.
Zein.Kallas@upc.edu
METHODOLOGICAL APPROACH: The second step the sensory test
 Where from we GET THE ENRICHED BEEF?
WE PRODUCE IT
 Cows were FED A DIET that include flax among other grains and
later TESTED IN LABORATORY for the content of Omega 3 and
LCA.
 ANIMALS WERE SLAUGHTERED with an average live weight of
458.4±16.6 kg at an EU-licensed commercial abattoir following
STANDARD PROCEDURES.
Zein.Kallas@upc.edu
Results: SENSORY TEST
Overall acceptability
Without
With
information
information
5.73
5.70
6.17
6.10
6.04
5.76
5.74
5.79
Type of beef meat
Conventional
Enriched with omega-3
Enriched with CLA
Enriched with omega-3 & CLA
P value**
0.858
0.611
0.051
0.712
1. INFORMATION about the benefits of n-3 and CLA fatty acids
did
NOT
HAVE
AN
INFLUENCE
on
their
SENSORY
EVALUATIONS for (P>0.05).
2. Beef ENRICHED with n-3 fatty acids had HIGHER LIKING
scores than beef from the other three treatments.
Zein.Kallas@upc.edu
Results: Impact of INFORMATION on the relative importance
Generalized Multinomial Logit model (Relative importance)
Attributes
Fat content
Colour
Origin
Diet
Price
Without information
Pre Sensory
Post sensory
With information
Pre Sensory
Post sensory
36.74***
19.46%***
19.62**
16.45%***
(25.1; 48.4)
(12.2; 26.7)
(7.7; 31.6)
(9.4; 23.5)
21.88***
11.78%***
16.92***
10.57%***
(13.1; 30.6)
(7.2; 16.2)
(10.3; 23.6)
(6.5; 14.6)
19.34***
7.14%***
15.07***
8.62%***
(12.7; 25.9)
(3.2; 11.0)
(9.1; 21.0)
(4.8; 12.4)
1.72
35.98%***
22.73***
34.79%***
(-7.6; 11.0)
(26.9; 45.1)
(15.4; 30.0)
(25.8; 43.8)
20.33***
25.66%***
25.67***
29.57%***
(11.1; 29.5)
(19.3; 31.9)
(17.4; 33.9)
(23.4; 35.8)
 The ORDER OF THE RELATIVE IMPORTANCE of the expected
preference was slightly different, showing that the information
provided HAD AN IMPACT on their choices.
Zein.Kallas@upc.edu
Results: Impact of INFORMATION on the relative importance
Generalized Multinomial Logit model (Relative importance)
Attributes
Fat content
Colour
Origin
Diet
Price
Without information
Pre Sensory
Post sensory
With information
Pre Sensory
Post sensory
36.74***
19.46%***
19.62**
16.45%***
(25.1; 48.4)
(12.2; 26.7)
(7.7; 31.6)
(9.4; 23.5)
21.88***
11.78%***
16.92***
10.57%***
(13.1; 30.6)
(7.2; 16.2)
(10.3; 23.6)
(6.5; 14.6)
19.34***
7.14%***
15.07***
8.62%***
(12.7; 25.9)
(3.2; 11.0)
(9.1; 21.0)
(4.8; 12.4)
1.72
35.98%***
22.73***
34.79%***
(-7.6; 11.0)
(26.9; 45.1)
(15.4; 30.0)
(25.8; 43.8)
20.33***
25.66%***
25.67***
29.57%***
(11.1; 29.5)
(19.3; 31.9)
(17.4; 33.9)
(23.4; 35.8)
 FAT CONTENT was the MOST IMPORTANT attribute for the
UNINFORMED consumers while it was LESS IMPORTANT for
INFORMED one. There is a clear SUBSTITUTION EFFECT
between the DIET and the FAT CONTENT.
Zein.Kallas@upc.edu
Results: Impact of INFORMATION on the relative importance
Generalized Multinomial Logit model (Relative importance)
Attributes
Fat content
Colour
Origin
Diet
Price
Without information
Pre Sensory
Post sensory
With information
Pre Sensory
Post sensory
36.74***
19.46%***
19.62**
16.45%***
(25.1; 48.4)
(12.2; 26.7)
(7.7; 31.6)
(9.4; 23.5)
21.88***
11.78%***
16.92***
10.57%***
(13.1; 30.6)
(7.2; 16.2)
(10.3; 23.6)
(6.5; 14.6)
19.34***
7.14%***
15.07***
8.62%***
(12.7; 25.9)
(3.2; 11.0)
(9.1; 21.0)
(4.8; 12.4)
1.72
35.98%***
22.73***
34.79%***
(-7.6; 11.0)
(26.9; 45.1)
(15.4; 30.0)
(25.8; 43.8)
20.33***
25.66%***
25.67***
29.57%***
(11.1; 29.5)
(19.3; 31.9)
(17.4; 33.9)
(23.4; 35.8)
 It is evident that consumers are LESS CONCERNED about the
AMOUNT OF VISIBLE FAT in beef, as long as it IS ENRICHED
with beneficial fatty acids.
Zein.Kallas@upc.edu
Results: Impact of SENSORY experience on the relative importance
Generalized Multinomial Logit model (Relative importance)
Attributes
Fat content
Colour
Origin
Diet
Price
Without information
Pre Sensory
Post sensory
With information
Pre Sensory
Post sensory
36.74***
19.46%***
19.62**
16.45%***
(25.1; 48.4)
(12.2; 26.7)
(7.7; 31.6)
(9.4; 23.5)
21.88***
11.78%***
16.92***
10.57%***
(13.1; 30.6)
(7.2; 16.2)
(10.3; 23.6)
(6.5; 14.6)
19.34***
7.14%***
15.07***
8.62%***
(12.7; 25.9)
(3.2; 11.0)
(9.1; 21.0)
(4.8; 12.4)
1.72
35.98%***
22.73***
34.79%***
(-7.6; 11.0)
(26.9; 45.1)
(15.4; 30.0)
(25.8; 43.8)
20.33***
25.66%***
25.67***
29.57%***
(11.1; 29.5)
(19.3; 31.9)
(17.4; 33.9)
(23.4; 35.8)
 Significant MODIFICATIONS for the UNINFORMED consumers,
while MINOR CHANGES occurred for INFORMED one.
Zein.Kallas@upc.edu
Results: Impact of SENSORY experience on the relative importance
Generalized Multinomial Logit model (Relative importance)
Attributes
Fat content
Colour
Origin
Diet
Without information
Pre Sensory
Post sensory
With information
Pre Sensory
Post sensory
36.74***
19.46%***
19.62**
16.45%***
(25.1; 48.4)
(12.2; 26.7)
(7.7; 31.6)
(9.4; 23.5)
21.88***
11.78%***
16.92***
10.57%***
(13.1; 30.6)
(7.2; 16.2)
(10.3; 23.6)
(6.5; 14.6)
19.34***
7.14%***
15.07***
8.62%***
(12.7; 25.9)
(3.2; 11.0)
(9.1; 21.0)
(4.8; 12.4)
1.72
35.98%***
22.73***
34.79%***
(-7.6; 11.0)
(26.9; 45.1)
(15.4; 30.0)
(25.8; 43.8)
20.33***
25.66%***
25.67***
29.57%***
Price

Consumers (11.1;
with
the ABSENCE
OF INFORMATION,
may have
29.5)
(19.3; 31.9)
(17.4; 33.9)
(23.4; 35.8)
TENTATIVE AND UNCERTAIN EXPECTATIONS. Therefore, the
significant changes between the EXPECTED and EXPERIENCED
preferences ARE MORE REMARKABLE.
Zein.Kallas@upc.edu
Results: Impact of SENSORY experience on the relative importance
Generalized Multinomial Logit model (Relative importance)
Attributes
Fat content
Colour
Origin
Diet
Price
Without information
Pre Sensory
Post sensory
With information
Pre Sensory
Post sensory
36.74***
19.46%***
19.62**
16.45%***
(25.1; 48.4)
(12.2; 26.7)
(7.7; 31.6)
(9.4; 23.5)
21.88***
11.78%***
16.92***
10.57%***
(13.1; 30.6)
(7.2; 16.2)
(10.3; 23.6)
(6.5; 14.6)
19.34***
7.14%***
15.07***
8.62%***
(12.7; 25.9)
(3.2; 11.0)
(9.1; 21.0)
(4.8; 12.4)
1.72
35.98%***
22.73***
34.79%***
(-7.6; 11.0)
(26.9; 45.1)
(15.4; 30.0)
(25.8; 43.8)
20.33***
25.66%***
25.67***
29.57%***
(11.1; 29.5)
(19.3; 31.9)
(17.4; 33.9)
(23.4; 35.8)
 The RELATIVE IMPORTANCE of the DIET attribute INCREASED
significantly once the products were EXPERIENCED and turned
to be one of the most important factors.
Zein.Kallas@upc.edu
Without information
With information
Expected
Experienced
Expected
Experienced
Random Parameters in utility functions ()
Moderate visible fat
-0.50***
-0.41***
-0.31***
-0.35***
Pale red
-0.30***
-0.25***
-0.27***
-0.22***
Other Spanish origin
-0.26**
-0.15**
-0.24***
-0.18***
Enriched with n-3
0.03
0.58***
0.19***
0.42***
Enriched with CLA
-0.01
0.12
0.01
0.10
*
***
Enriched with n-3 & CLA
0.11
0.22
0.33
0.49***
Price 6.6€ (high)
-0.52***
-0.69***
-0.67***
-0.87***
Price 5.7€ (medium-high)
0.11***
-0.07
0.20***
0.02
Price 4.8€ (medium-low)
0.37***
0.38***
0.32***
0.41***
Opt-Out
0.48***
0.44***
-0.15
0.16
Scale parameters
Variance parameter
in scale
0.56***on
Impact
ofparameter
INFORMATION
Weighting parameter Gamma
0.34***
0.08
0.12*
expected
preferences
0.31***
0.10
0.11
0.10
Standard deviations of parameters distribution
1.63***
1.58***
2.14***
1.82***
0.95***
0.11**
0.90***
0.19***
0.15**
0.26***
0.16***
0.28***
0.48***
1.08***
0.52***
1.71***
0.31***
1.37***
0.20
0.52***
0.56***
1.97***
0.65***
2.099***
1.67***
1.48***
1.41***
1.57***
0.61***
0.76***
0.51***
0.76***
1.06***
0.916***
0.77***
0.79***
2.39***
2.30***
2.12***
2.60***
-2,658.67
-2705.29
-2656.7667
-2801.51
-3,571.09
-3,571.09
-3,604.36
-3,604.36
***
***
***
1,824.84
1,731.60
1,895.19
1605.69***
0.255
0.242
0.262
0.227
2.124
2.160
2.103
2.214
Std. Dev. Moderate
visible
fatNOT RECEIVE INFORMATION about the benefits
 Consumers
that
DO
Std. Dev. Pale red
Std. Dev. Other Spanish origin
Std. Dev. Enriched with n-3
Std. Dev. Enriched with CLA
Std. Dev. Enriched with n-3 & CLA
Std. Dev. Price 6.6€ (high)
Std. Dev. Price 5.7€ (medium-high)
Std. Dev. Price 4.8€ (medium-low)
Std. Dev. Opt-Out
Log-Likelihood (θ)
Log-Likelihood (0)
LL ratio test
Pseudo R2
AIC/N
of n-3 and CLA fatty acids or their role in human health are more
CONCERNED about the AMOUNT of FAT in meat.
Without information
With information
Expected
Experienced
Expected
Experienced
Random Parameters in utility functions ()
Moderate visible fat
-0.50***
-0.41***
-0.31***
-0.35***
Pale red
-0.30***
-0.25***
-0.27***
-0.22***
Other Spanish origin
-0.26**
-0.15**
-0.24***
-0.18***
Enriched with n-3
0.03
0.58***
0.19***
0.42***
Enriched with CLA
-0.01
0.12
0.01
0.10
*
***
Enriched with n-3 & CLA
0.11
0.22
0.33
0.49***
Price 6.6€ (high)
-0.52***
-0.69***
-0.67***
-0.87***
Price 5.7€ (medium-high)
0.11***
-0.07
0.20***
0.02
Price Impact
4.8€ (medium-low)
0.37***
0.38***
0.32***
0.41***
of sensory experience
on
expected
preferences
Opt-Out
0.48***
0.44***
-0.15
0.16
Scale parameters
Variance parameter in scale parameter
0.56***
0.08
0.12*
 The
utility
for the
ENRICHED
WITH
***
Weighting
parameter
Gamma
0.34MEAT
0.31***N-3 increased.
0.10
0.11
0.10
Standard deviations of parameters distribution
1.82***
0.95***
0.11**
0.90***
0.19***
0.15**
0.26***
0.16***
0.28***
0.48***
1.08***
0.52***
1.71***
0.31***
1.37***
0.20
0.52***
0.56***
1.97***
0.65***
2.099***
1.67***
1.48***
1.41***
1.57***
0.61***
0.76***
0.51***
0.76***
1.06***
0.916***
0.77***
0.79***
2.39***
2.30***
2.12***
2.60***
-2,658.67
-2705.29
-2656.7667
-2801.51
-3,571.09
-3,571.09
-3,604.36
-3,604.36
***
***
***
1,824.84
1,731.60
1,895.19
1605.69***
0.255
0.242
0.262
0.227
2.124
2.160
2.103
2.214
 The
of preference
for 1.63
the***CLA ENRICHED
BEEF.
Std. LACK
Dev. Moderate
visible fat
1.58***
2.14***
Std. Dev. Pale red
Std. Dev. Other Spanish origin
Std. Dev. Enriched with n-3
Std. Dev. Enriched with CLA
Std. Dev. Enriched with n-3 & CLA
Std. Dev. Price 6.6€ (high)
Std. Dev. Price 5.7€ (medium-high)
Std. Dev. Price 4.8€ (medium-low)
Std. Dev. Opt-Out
Log-Likelihood (θ)
Log-Likelihood (0)
LL ratio test
Pseudo R2
AIC/N
 Consumers may be LESS CONCERNED about the AMOUNT OF
FAT in meat if they are aware that SENSORY PROPERTIES are
NOT COMPROMISED when meat is enriched with BENEFICIAL
FATTY ACIDS.
Without information
With information
Expected
Experienced
Expected
Experienced
Random Parameters in utility functions ()
Moderate visible fat
-0.50***
-0.41***
-0.31***
-0.35***
Pale red
-0.30***
-0.25***
-0.27***
-0.22***
Other Spanish origin
-0.26**
-0.15**
-0.24***
-0.18***
Enriched with n-3
0.03
0.58***
0.19***
0.42***
Enriched with CLA
-0.01
0.12
0.01
0.10
*
***
Enriched with n-3 & CLA
0.11
0.22
0.33
0.49***
Price 6.6€ (high)
-0.52***
-0.69***
-0.67***
-0.87***
Price 5.7€ (medium-high)
0.11***
-0.07
0.20***
0.02
Price 4.8€ (medium-low)
0.37***
0.38***
0.32***
0.41***
Opt-Out
0.48***
0.44***
-0.15
0.16
Scale parameters
Variance parameter in scale parameter
0.56***
0.08
0.12*
0.11
Weighting parameter Gamma
0.34***
0.31***
0.10
0.10
Standard deviations of parameters distribution
1.63***
1.58***
2.14***
1.82***
0.95***
0.11**
0.90***
0.19***
0.15**
0.26***
0.16***
0.28***
0.48***
1.08***
0.52***
1.71***
0.31***
1.37***
0.20
0.52***
0.56***
1.97***
0.65***
2.099***
1.67***
1.48***
1.41***
1.57***
0.61***
0.76***
0.51***
0.76***
1.06***
0.916***
0.77***
0.79***
2.39***
2.30***
2.12***
2.60***
-2,658.67
-2705.29
-2656.7667
-2801.51
-3,571.09
-3,571.09
-3,604.36
-3,604.36
***
***
***
1,824.84
1,731.60
1,895.19
1605.69***
0.255
0.242
0.262
0.227
2.124
2.160
2.103
2.214
Std. Dev. Moderate
 Focusing
on visible
the fat
SCALE parameters, the INFORMATION and the
Std. Dev. Pale red
Std. Dev. Other Spanish origin
Std. Dev. Enriched with n-3
Std. Dev. Enriched with CLA
Std. Dev. Enriched with n-3 & CLA
Std. Dev. Price 6.6€ (high)
Std. Dev. Price 5.7€ (medium-high)
Std. Dev. Price 4.8€ (medium-low)
Std. Dev. Opt-Out
Log-Likelihood (θ)
Log-Likelihood (0)
LL ratio test
Pseudo R2
AIC/N
SENSORY EXPERIENCE have had a significant impact.
 The average error scales DECREASED when moving from for
UNINFORMED TO INFORMED consumers.
 This may indicate that when the CONSUMERS are INFORMED they
make MORE RELIABLE choices.
Without information
With information
Expected
Experienced
Expected
Experienced
Random Parameters in utility functions ()
Moderate visible fat
-0.50***
-0.41***
-0.31***
-0.35***
Pale red
-0.30***
-0.25***
-0.27***
-0.22***
Other Spanish origin
-0.26**
-0.15**
-0.24***
-0.18***
Enriched with n-3
0.03
0.58***
0.19***
0.42***
Enriched with CLA
-0.01
0.12
0.01
0.10
*
***
Enriched with n-3 & CLA
0.11
0.22
0.33
0.49***
Price 6.6€ (high)
-0.52***
-0.69***
-0.67***
-0.87***
Price 5.7€ (medium-high)
0.11***
-0.07
0.20***
0.02
Price 4.8€ (medium-low)
0.37***
0.38***
0.32***
0.41***
Opt-Out
0.48***
0.44***
-0.15
0.16
Scale parameters
Variance parameter in scale parameter
0.56***
0.08
0.12*
0.11
Weighting parameter Gamma
0.34***
0.31***
0.10
0.10
Std. Dev. Moderate visible fat
Std. Dev. Pale red
Std. Dev. Other Spanish origin
Std. Dev. Enriched with n-3
Std. Dev. Enriched with CLA
Std. Dev. Enriched with n-3 & CLA
Std. Dev. Price 6.6€ (high)
Std. Dev. Price 5.7€ (medium-high)
Std. Dev. Price 4.8€ (medium-low)
Std. Dev. Opt-Out
Log-Likelihood (θ)
Log-Likelihood (0)
LL ratio test
Pseudo R2
AIC/N
Standard deviations of parameters distribution
1.63***
1.58***
2.14***
1.82***
0.95***
0.11**
0.90***
0.19***
0.15**
0.26***
0.16***
0.28***
0.48***
1.08***
0.52***
1.71***
0.31***
1.37***
0.20
0.52***
0.56***
1.97***
0.65***
2.099***
1.67***
1.48***
1.41***
1.57***
0.61***
0.76***
0.51***
0.76***
1.06***
0.916***
0.77***
0.79***
2.39***
2.30***
2.12***
2.60***
-2,658.67
-2705.29
-2656.7667
-2801.51
-3,571.09
-3,571.09
-3,604.36
-3,604.36
***
***
***
1,824.84
1,731.60
1,895.19
1605.69***
0.255
0.242
0.262
0.227
2.124
2.160
2.103
2.214
 SENSORY experience showed EVIDENCE of a SHIFT in the scaling
factor across choice (the error scale turns out to be insignificant).
 After the TASTING EXPERIENCE, consumers tended to make more
RELIABLE CHOICE.
Conclusions
 Providing INFORMATION to consumers about the role of
beneficial fatty acids and their potential HEALTH BENEFITS
would FAVOR MARKETING of n-3 enriched beef through
modifications in animal diet.
 SENSORY EXPERIENCE had a significant IMPACT on defining
consumer preferences for beef attributes, especially for
UNINFORMED consumer.
 Enrichment of beef with CLA was NOT POSITIVELY valued
by consumers, REGARDLESS of the INFORMATION provided
and the TASTING experience.
Zein.Kallas@upc.edu
conclusions
 The, SENSORY TEST and the INFORMATION DECREASED the
SCALE HETEROGENEITY term.
 MARKETING OPPORTUNITIES for n-3 enriched beef seem
PROMISING in SPAIN.
 Analyzing the attributes NON-ATTENDANCE BEFORE AND
AFTER SENSORY are proposed for further research
Zein.Kallas@upc.edu
Thank you for
your attention
Zein.Kallas@upc.edu