Applied Food Research 2 (2022) 100038

Contents lists available at ScienceDirect

Applied Food Research

journal homepage: www.elsevier.com/locate/afres

Sensory profile, consumer acceptability and preference mapping of

cassava-chia seeds composite porridges
Ashura Katunzi-Kilewela a,b,∗, Richard J. Mongi Dr c, Lillian Daniel Kaale Dr a, Oscar Kibazohi
Dr a, Roman Mmanda Fortunatus a, Leonard MP Rweyemamu Dr a
a
Department of Food Science and Technology, University of Dar es Salaam, P.O. Box 35134 Dar es Salaam, Tanzania
b
Tanzania Bureau of Standards, P.O Box 9524 Dar es Salaam, Tanzania.
c
Department of Public Health, University of Dodoma, P.O Box 395 Dodoma Tanzania, Tanzania

a r t i c l e i n f o a b s t r a c t

Keywords: This study applied quantitative descriptive analysis, hedonic test, and preference mapping techniques to inves-
Quantitative descriptive analysis tigate the sensory profile, consumer acceptability, and drivers for consumer acceptability of cassava-chia seeds
Consumer test composite porridge samples. Chia seeds flour was developed and blended with cassava flour at varying propor-
Porridge
tions (0, 5, 10, 15, 20, and 25%) to make six composite porridges labelled CF, CCSF1, CCSF2, CCSF3, CCSF4,
Cassava flour
and CCSF5. The control cassava porridge (CF) formulation had significantly (p <0.05) higher sweetness and
Chia seeds flour
aroma intensities than the composite formulations. Conversely, the CCSF4 (20%) and CCSF5 (25%) composite
formulations had significantly (p < 0.05) higher colour, mouthfeel, and oiliness intensities than the control for-
mulation (CF). Furthermore, when compared to cassava porridge, the overall acceptability of composite porridge
samples was significantly (p <0.05) much greater. Consumer preference for CCSF5 was much greater than for
CF, although sample CCSF5 did not differ significantly (p > 0.05) from other composite porridge blends (sample
CCSF1, CCSF3, and CCSF4). Consumer preferences for porridge formulations were largely driven by colour, tex-
ture, and oiliness attributes. Thus, mixing up to 25% chia seeds flour with cassava flour produces a composite
porridge with enhanced sensory properties and, hence, more consumer appeal than non-mixed cassava porridge.
Significantly, this study creates an opportunity for small to large-scale food processors to valorise cassava flour
by producing cassava-chia seeds composite flour to benefit the broader population in terms of food security and
nutritional enhancement.

1. Introduction quently serve as materials for processing food products such as porridge
and/or hardened porridge (Muoki, 2013), commonly known as ugali or
Porridge is generally a cereal-based liquid-to-semi-solid food product cassava meal in East Africa.
with a soft texture that all age groups can consume, especially children However, cassava contains low protein content (1-2%)
and patients, due to its smoothness and easy ingestion (Alyami et al., (Bayata, 2019; Laswai, Silayo, Mpagalile, Balegu, & John, 2006),
2019). The porridge serves as a weaning food for babies and an everyday which is far too little to meet the recommended dietary allowance
breakfast meal in schools and households for adults (Ajifolokun, Bas- (RDA) for infants aged 0 - 4 months, which is 8g per day and infants
son, Osunsanmi, & Zharare, 2019). Characteristically, porridge has high aged 4 - 12 months (11g per day) (Richter et al., 2019), respectively.
starch content, but less flour-to-water ratio (5-15% w/v) due to the for- Infants aged 0 - 4 months and 4 -12 months need at least 400 and 550g
mation of viscous liquid and is less nutritious due to its composition per day of cassava intake to meet the basic protein requirement, which
(Taylor & Emmambux, 2008). Yet, cassava is one of the most common on the surface appears a lot for their age but which underlines the
starch-based and staple foods consumed in sub-Sahara Africa after maize necessity for more nutrient-dense food formulation. Standard practices
and rice (Muoki, 2013). The crop is famous for being drought-resistant, in the East African region aimed to enhance energy density and nutri-
easy processing into different forms, and for playing a crucial role in tional composition of local foods include the formulation of composite
fostering food security in many developing countries (Alphonce, Kaale, flour with nutrient-dense foods (Onyango, Luvitaa, Unbehend, & Haase,
Millinga, & Rweyemamu, 2020). Various forms of cassava flour fre- 2020).

∗
Corresponding author.
E-mail addresses: [email protected] (A. Katunzi-Kilewela), [email protected] (R.J. Mongi Dr), [email protected] (L.D. Kaale Dr),
[email protected] (O. Kibazohi Dr), [email protected] (R.M. Fortunatus), stayfi[email protected] (L.M. Rweyemamu Dr).

https://doi.org/10.1016/j.afres.2021.100038
Received 11 July 2021; Received in revised form 2 December 2021; Accepted 22 December 2021
Available online 26 December 2021
2772-5022/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

A study in Rwanda exposed the inadequacy of nutrients in most Table 1

locally-marketed porridge-type complementary foods in terms of pro- Cassava-chia seeds flour bends formulations.
tein, fat, and vitamins A & E (Grosshagauer et al., 2020). The main in- Sample type Dried mashed cassava (g) Chia seed grains (g)
gredients were cereal-based foods such as maize, sorghum, rice, and
Control (CF) 1000 0
millet, which are deficient in micronutrients and essential amino acids
CCSF1 950 50
such as lysine and tryptophan (Nout & Rombouts, 1992). The compar- CCSF2 900 100
ison of these complementary foods nutrient requirements was bench- CCSF3 850 150
marked based on Guidelines for formulated complementary foods de- CCSF4 800 200
CCSF5 750 250
signed for older infants and young children (FAO/WHO, 2013) and the
World Food Programme (2015) (Grosshagauer et al. (2020). Implicitly, Each composite flour formulated weighed 1 ± 0.01 kg. Each
there is a need to choose and blend locally-available nutritious ingre- 1kg composite flour was packed in moisture-proof zip lock low-
dients with tubers, grains, or legumes to meet the recommended daily density polyethylene bags (LDPE) for laboratory use and stored
nutritional requirements (Bayata, 2019; Laswai et al., 2006). at ambient temperature before laboratory analysis.
Chia seeds have become a common plant grown in East Africa, in-
cluding Tanzania, due to their high nutritional composition. The seeds 2.2. Preparation of cassava-chia seeds composite flour blends
can be blended with cassava to make cassava-chia seeds composite
flour at the household level. Chia seeds reportedly have protein (15 - Cassava tubers were peeled, washed with portable clean water,
25%), fats (30 - 33%), ash (4 - 5%), dietary fibre (18 - 30%), miner- grated, and pressed to reduce water content. The grated and pressed
als, vitamins, antioxidants, and other phenolic compounds (Coelho & cassava mash was dried using a solar drier at 25 – 55°C for 8 hours
Salas-Mellado, 2014; Marcinek & Krejpcio, 2017). Chia seeds increased (Aso, Teixeira, & Welt, 2015). Chia seeds were cleaned to remove de-
nutritional contents (protein, fat, ash) in cassava-chia seeds extruded bris, stones, and suspended extraneous matter. The dried cassava and
flour (Otondi, Nduko, & Omwamba, 2020). However, the detailed effect cleaned chia seeds were mixed to make cassava-chia seeds composite
of un-extruded flour blends sensory attributes in porridge made from flour of five different mixing ratios, as indicated in
blends remains largely unknown. Yet, replacing chia seeds in wheat Table 1. Chia seeds amount varied from 50 – 250 g to produce 1 kg
bread up to 300g per kg enhanced nutritional benefits without nec- of composite flour to meet the recommended dietary intake (RDA) of
essarily compromising the overall acceptability of the bread (Zhu & protein for children aged between 0 and 10 (Richter et al., 2019) while
Chan, 2018). Moreover, the blending of cassava and chia seeds can re- considering the impact on the sensory characteristics of the composite
sult in different sensory characteristics due to the interaction between blends relative to cassava flour. The mixed ingredients of dried mashed
food components. Porridge viscosity, colour, taste, mouthfeel, aroma are cassava and chia seeds were then milled together using a laboratory
common sensory attributes that affect consumer acceptance and prefer- blender (Waring Commercial, USA) and sieved to 0.2 mm mesh size
ence of the product (Onyango et al., 2020). to obtain uniformly-sized particle distribution for each composite flour
Sensory evaluation is an essential aspect of food product develop- blend. Whole cassava flour (CF) served as a control sample, and blended
ment. The quantitative descriptive analysis technique provides a quan- cassava-chia seeds flour (CCSF) of five different combinations served as
titative description of a food product based on its sensory properties test samples (Table 1).
(Lawless & Heymann, 2010b). In this regard, the consumer test ascer-
tains whether the consumers prefer or accept the product based on 2.3. Porridge preparation
its sensory characteristics (Lawless & Heymann, 2010b). Additionally,
preference mapping is a multivariate analysis method relating sensory Six different porridges were prepared from cassava flour (control)
and acceptability data and informs how the attributes drive consumer and five composite flour blends using a modified method described by
preference (Yenket, 2011). It is a valuable tool for gaining a deeper Emmaculate, Wandayi, Ooko, and Kadenyeka (2020). For each sample,
understanding of consumer preferences (Ubeda, Callejón, Troncoso, & 100g of flour formulation was weighed using laboratory electronic ana-
Morales, 2017). And yet, the literature on the sensory profile, consumer lytical balance (Mettler XP205 Toledo, United States), then mixed with
acceptability, and preference mapping of cassava-chia seeds porridge re- 300mL tap water to make a semi-liquid mixture, and then added to
mains largely lacking. This study, therefore, was conducted to establish a stainless-steel pot containing 700 mL of boiled water on an electric
and avail the missing information. cooker set at 120°C. The mixture was stirred continuously for 5 min
using a flat wooden spoon until the porridge formed. The stirring of
the porridge continued and left to boil for another 5 min. Before analy-
sis, the six porridge formulations were then poured into thermos 3.2 L
thermos. The porridge temperature was kept constant at 50°C using the
2. Materials and methods thermos before being served to the panellists.

2.1. Sample collection 2.3.1. Quantitative Descriptive Analysis (QDA)

2.3.1.1. Selection and training of panellist. Descriptive sensory profiling
Cassava roots (Kiroba variety) were bought from a farmer in Bag- was conducted at the Department of Food Science and Technology of the
amoyo district, Coast (Pwani) region who sourced cassava cultivars from University of Dar es Salaam. The panellists aged 23 - 28 years were Food
Mikocheni Agricultural Research Institute at Chambezi Centre in Coast Science and Technology graduates of the 2020 cohort at the University
region, Tanzania. This study used this variety because it has the most sig- of Dar es Salaam. Nineteen (19) panellists drawn from 36 students were
nificant potential for producing cassava roots with high starch and less selected to participate in the study. The training was conducted for five
cyanide content. Furthermore, it is a preferable cultivar for cassava flour days following the ISO Standard (1993). The selection criteria were the
production commonly cultivated on coast plains (Coast region, Tanga, panellists’ commitment, motivation, good health, and ability to work
Lindi, and Mtwara) in Tanzania with an average of 9 - 12 months mat- in a team (Lawless & Heymann, 2010a). During the training, panellists
uration until cassava roots are ready for harvesting (Mtunguja et al., developed attributes, and a quantitative scale for attribute scoring was
2019). The harvested cassava tubers were then transported to the De- provided. A range of products such as cassava porridge, flour, tomato
partment of Food Science of the University of Dar es Salaam for further ketchup, cooking oil, and mango pudding was provided as reference
pre-laboratory analysis processing. Chia seeds were bought from a local materials to facilitate the process. Out of the numerous attributes de-
farmer in the Karagwe district, Kagera region. veloped, only viscosity, colour, taste, mouthfeel, jellying, oiliness, and

2
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

Table 2
Sensory descriptors panellists used to evaluate cassava-chia seeds porridge.

Attribute Description Reference Scale ranges (1-9)

Viscosity The resistance of the Canadian harvest Less viscous - 1

product to flow during Tomato ketchup More viscous - 9
consumption (from the
cup to the mouth)
Colour Characteristic of visual Khaki brown Low intensity - 1
perception described High intensity - 9
through colour
associated with khaki
brown
Taste The taste associated with Cooked cassava Not concentrated - 1
a cooked cassava porridge Very concentrated -
porridge 9
Mouthfeel The degree to which the Millet flour Fine particles - 1
particle size of the Coarse particles - 9
product is felt in the
mouth
Aroma Aroma associated with Dried cassava aroma Less aromatic - 1
dried cassava aroma Very aromatic - 9
Jellying The visual appearance of Set mango pudding Not jelly-like -1
the jelly-like behaviour Jelly like - 9
of the product
Oiliness The degree to which the Sunflower cooking Less oily - 1
product is felt oil when oil Very oily - 9
scrubbed between two
fingers
Source: Study QDA panellists (2020).

aroma were agreed upon and selected for the study, as Table 2 demon- observed (Amerine et al., 2013; Civille & Carr, 2015; Lawless & Hey-
strates. Furthermore, the panellists developed a nine-point quantitative mann, 2010b, 2013).
line scale for attributes intensity, scoring with one for the lowest and
nine for the highest. 2.4.1.4. Preference test. The preference ranking test was carried out in
the Department of Food Science Laboratory by 85 lay consumers. The
consumers were drawn from UDSM students and staff and TBS person-
2.3.1.2. Panel performance evaluation. Before the actual test, the panel’s
nel aged 18 and above years using a 5-point ranking scale (whereby
performance in terms of agreement and ability of individual panellists to
1 = most preferred and 5 = least preferred) as Lawless and Hey-
discriminate between samples and reproducibility was assessed in two
mann (2013) described. Consumers were provided with six porridge
sessions using the Panel Check software VI.4.2 (Tomic et al., 2010). A
samples, each coded with a new three-digit random number. The tasters
small portion of the porridge samples was coded with 3-digit random
were then asked to rank their preferences using numbers indicated on
numbers and presented randomly to each panellist at 11:00 hrs and
a scale. Good sensory practices involving testing area, sample coding,
asked to rate the attribute’s intensity using the scale provided. Water
panellist, and testing protocol such as rinsing the mouth between tests
was provided for rinsing the mouth between the tests during the anal-
and drinking water were also observed (Amerine et al., 2013; Civille &
ysis. Data subjected to univariate and multivariate analyses for Tucker
Carr, 2015; Lawless & Heymann, 2010b, 2013).
and pMSE plots, whereby 18 out of 19 panellists were retained for the
final actual descriptive test.
2.5. Statistical Data analysis

2.4. Sensory analysis Quantitative descriptive analysis (QDA) and hedonic data analy-
sis were carried out using the R statistical package (R Development
2.4.1. Consumer test Core Team, Version 3.0.0 Vienna, Austria) for the analysis of variance
2.4.1.3. Acceptability test. Eighty-five (85) untrained consumers drawn (ANOVA). The means were separated by Tukey’s Honestly Significant
from the UDSM students and staff and Tanzania Bureau of Standards Difference (THSD) test at p < 0.05. Non-parametric Friedman rank-sum
(TBS) personnel at the Department of Food Science Laboratory aged test was performed for preference data. Furthermore, Latentix Software
18 and above using a 9-point hedonic scale (whereby 1 = dislike ex- (Latentix Aps Team, version 2.12, Frederiksberg Denmark) was used to
tremely and 9 = like extremely) as described by Amerine, Pangborn, evaluate systematic variation in sensory data using the Principal Com-
and Roessler (2013), Civille and Carr (2015), and Lawless and Heymann ponent Analysis (PCA). Partial Least Square Regression was computed
(2010b, 2013). Subsequently, 50 mL of six different porridge samples using Panel Check software VI.4.2 (Tomic et al., 2010) to determine the
were served to each consumer at a temperature of 40 - 45°C in a ran- relationship between sensory and hedonic data and ascertain drivers
domised order coded with 3-digit random numbers at around 10:00 hrs. behind consumer liking (Martens & Martens, 2001). Finally, the results
The judges rated the degree of liking the taste, aroma, colour, texture, are presented as the arithmetic mean and standard deviation in tables,
mouthfeel, and overall acceptability attributes, indicating their degree graphs, and PCA plots.
of liking or disliking by putting a number provided in the hedonic scale
based on their preference. The paper ballots provided facilitated the in- 3. Results and discussion
dication of the degree of liking based on the definition of each attribute
used. Testing completed in one session and all the acceptable good sen- 3.1. Quantitative descriptive analysis of porridge samples
sory practices involving the testing area, sample coding, and serving
conditions such as constant temperature maintenance, panellists, and Mean intensity ratings of descriptive attributes between porridge
testing protocol, for example, rinsing the mouth between the tests were samples appear in Table 3. All the samples showed statistical variations

3
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

Table 3
Mean intensity scores of cassava-chia seeds composite flour porridge samples.

Sample Viscosity Colour Sweetness Mouthfeel Aroma Jellying Oiliness

a d a d a a
CF 6.4±1.38 2.7±1.35 7.4±2.71 2.3±0.99 8.3±0.83 6.3±1.94 4.4±2.53b
CCSF1 (95:5) 6.3±1.56a 6.0±1.21c 6.1±2.42b 3.7±1.80c 7.0±1.18b 5.9±1.71ab 5.1±2.39ab
CCSF2 (90:10) 6.3±1.65a 6.6±1.10b 5.8±2.38bc 4.1±1.73bc 6.4±1.38bc 5.7±1.88ab 5.5±1.86a
CCSF3 (85:15) 6.3±1.36a 7.6±1.11a 5.4±2.54c 4.9±1.84ab 6.0±1.65cd 5.5±1.81ab 5.4±1.90a
CCSF4 (80:20) 6.4±1.46a 7.9±1.07a 5.4±2.49c 5.2±2.18a 6.0±1.57cd 5.9±1.79ab 5.4±1.90a
CCSF5 (75:25) 6.6±1.42a 7.9±1.02a 5.5±2.27bc 5.2±2.21a 5.6±1.65d 5.6±1.90ab 5.4±1.84a

Results expressed as mean ± SD (n = 18). Mean values sharing different superscript letters across columns
significantly differ (p < 0.05). Key CF is cassava flour, and CCSF is cassava-chia seeds flour.

in mean intensity scores (p < 0.05) for all the attributes except viscos- Table 4
ity and jellying. Cassava porridge (CF) had the lowest colour, mouthfeel, Characteristics of Consumer Panel (n = 85).
and oiliness intensities significantly compared to all-composite porridge Variable Category Percentage (%)
blends with the highest intensities for the remaining attributes (Table 3).
Gender Male 48.2
Generally, the mean intensity scores of colour and mouthfeel tend to
Female 51.8
increase as the composition of chia seeds flour rises in the composite Age group 15-30 63.5
blends. On the other hand, the mean intensity scores for sweetness and 31-45 29.4
aroma dropped as the composition of chia seeds flour increased in the 46-60 7.1
Occupation of the Undergraduate 47.1
composite blends. No significant (p > 0.05) variations were observed
panellist Postgraduate 8.2
in mean intensity scores in colour, sweetness, mouthfeel, aroma, and Staff 44.7
oiliness intensities between samples CCSF3, CCSF4, and CCSF5.
Substitution of cassava flour with different levels of chia seeds flour Frequency of Daily 8.2
has a significant effect on the sensory profile of the final composite consumption of Once a week 42.4
porridge Once in a month 23.5
flour porridges. The progressive reduction in intensity scores for sweet-
Seldom 25.9
ness and aroma in composite flour porridges with increasing cassava
flour substitution with chia seeds flour could be associated with cas- Willingness to Yes 92.9
sava flour’s physiochemical characteristics and dilution. Sweet cassava purchase No 7.1
varieties naturally produce roots that taste sweet (Imakumbili, Semu,
Semoka, Abass, & Mkamilo, 2019). The substitution of cassava flour
with chia seeds flour in the composite porridge was observed to affect with cassava flour up to 25% enhances the intensity of the final compos-
the sensory profiles of the final product. Composite samples had cassava ite porridge’s sensory attributes. The intrinsic characteristics and com-
flour substituted up to 25% resulted in reduced sweetness and aroma position of raw materials used to formulate these composite flours are
compared to the control samples containing 100% cassava flour. responsible for this improvement.
Furthermore, chia seeds are characterised by tastelessness and black-
ness due to phenolic and other compounds (Marcinek & Krejpcio, 2017), 3.2. Consumer tests
which have no noticeable effect on porridge sweet flavour. The high
starch content in the porridge contributed to paste formation during 3.2.1. Consumer characteristics
cooking (Alamu, Maziya-Dixon, & Dixon, 2017). During heat treatment, The consumer panel comprised 85 panellists, 48.2% male and 51.8%
the cassava flour starch undergoes swelling and starts to gelatinise, female. Most of the consumer panel (63.5%) have an age group of 15
increasing the product’s viscosity (Tappiban et al., 2020). However, - 30 years. In addition, 47.1% were undergraduates, 8.2% were post-
heightened mixing of chia seeds with the composite flour reduced the graduates, and 44.7% were faculty members. Eight-point-two percent
starch content and, consequently, trimmed the composite porridges’ jel- (8.2%) of the panellists were daily porridge users, 42.4% used it once a
lying (pasting properties). This could be attributable to the presence week, 23.5% once per month, and 25.9% were seldom porridge users.
of high content of fat (30 – 33%) and protein (15 - 25%) in chia The findings show that 92.9% of the consumer panellists were willing
seeds (Ixtaina, Nolasco, & Tomas, 2008), which causes competition with to buy the products once introduced in the market, whereas a minority
starch molecules for hydration (Batey & Curtin, 2000). Also, increased (7.1%) were unwilling to buy the porridge (Table 4).
protein and fats in the composite porridge samples affect the swelling of
3.2.2. Acceptability test
starch due to the formation of food component’s complex structures and
The acceptability of different formulated porridge samples prepared
reduced viscosity (Pomeranz, 2012). Furthermore, chia seeds contain
using cassava-chia seeds flour blends, as shown in Table 5. There were
mucilage and gum, which contribute to high water absorption (Segura-
significant (p < 0.05) variations in the mean hedonic score in all the
Campos, Ciau-Solís, Rosado-Rubio, Chel-Guerrero, & Betancur-Ancona,
tested attributes except viscosity and mouthfeel. Control cassava por-
2014), hence less viscous effect.
ridge (CF) had the significantly (p > 0.05) lowest mean hedonic scores
The higher colour intensity in the composite porridge than the whole
for taste, colour, aroma, and overall acceptability compared to compos-
cassava porridge could be attributable to the blending and level of chia
ite porridge samples CCSF2 to CCSF5. The composite porridge samples
seeds grated in the formulations. The colour changes in the compos-
CCSF2 to CCSF5 were statistically similar (p > 0.05) for taste, aroma,
ite porridge are due to the incorporation of chia seeds which are natu-
and overall acceptability intensities and most acceptable samples. The
rally brown or black (Ixtaina et al., 2008; Muñoz, Cobos, Diaz, & Aguil-
findings show that consumer liking of the samples increased as the pro-
era, 2012). Their combination with cassava flour produced a khaki-
portion of chia seeds flour increased in the formulation.
brown colour whose intensity in the porridge samples increased with
the heightening level of cassava flour substitution. Moreover, the higher 3.3. Preference test
oiliness intensity in porridge samples at elevated cassava flour substitu-
tion levels is linked to chia seeds containing high-fat content (30 - 33%) Consumer preference for porridge samples is as presented in Fig. 1.
(Ixtaina et al., 2008). The results suggest that blending chia seeds flour Sample CCSF5 had significantly (p < 0.05) the lowest rank-sum com-

4
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

Table 5
Mean hedonic scores of porridge samples.

Sample Viscosity Taste Colour Mouthfeel Aroma Jellying Oiliness Overallacceptability

a c d a b b c
CF 7.1±1.48 6.0±1.90 4.2±1.72 6.6±1.91 6.3±1.93 6.2±1.73 5.8±1.75 6.4±1.79c
CCSF1 (95:5) 7.2±1.56a 6.2±1.90bc 5.8±1.54c 6.6±1.83a 6.5±1.64ab 6.2±1.74b 5.9±1.69c 6.7±1.62b
CCSF2 (90:10) 7.4±1.29a 6.8±1.69ab 6.5±1.43b 6.9±1.51a 6.6±1.68a 6.5±1.52ab 6.8±1.32ab 6.9±1.61ab
CCSF3 (85:15) 6.9±1.80a 6.8±1.70ab 6.6±1.92b 6.7±1.97a 6.8±1.91ab 7.0±1.34a 6.5±1.64b 7.0±1.39a
CCSF4 (80:20) 6.8±1.78a 6.8±1.66ab 6.8±1.37ab 6.7±1.75a 6.7±1.58ab 6.9±1.38a 6.9±1.35a 7.1±1.24a
CCSF5 (75:25) 6.9±1.36a 6.9±1.29a 7.3±1.47a 6.6±1.61a 7.1±1.44a 6.7±1.51ab 7.0±1.45a 7.1±1.32a

Results expressed as mean ± SD (n = 85). Mean values sharing different superscript letters across columns significantly differ (p <
0.05). Key CF is cassava flour, and CCSF is cassava-chia seeds flour.

Fig. 1. The rank-sum of porridge samples (n = 85). Bars with different letters significantly differ at p < 0.05. Least Significant Rank Difference (LSRD = 52).

pared to the highest value in sample CF. Based on the scale in use, CCSF5 3.4. Relationship between descriptive data and hedonic liking through
was the most preferred sample. However, sample CCSF5 was not signif- preference mapping
icantly different from other composite porridge blends (sample CCSF1,
CCSF3, and CCSF4). 3.4.1. Principal component of descriptive sensory data
These results suggest that consumer preference progressively rose Fig. 2 shows a bi-plot with the two first significant principal compo-
as the chia seeds proportion increased in the formulations. Indeed, the nents from Principal Component Analysis (PCA) about the average sen-
higher the consumer hedonic scores observed between CCSF5 and CF sory attributes of the porridge samples. Principal component 1 (PC 1)
porridges for taste, colour, and oiliness attributes, respectively, might accounted for 97.594% of the total variations, whereas PC 2 accounted
have been enhanced because of blending 25% chia seeds flour in the for 1.634%. Generally, the samples show the distribution in three PC1
composite formulation. The increased mixture of chia seeds in cas- groups: sample CCSF3 to CCSF5, CCSF1 to CCSF2, and CF. PC 1 con-
sava flour also reduced the after-taste flavour of the dried cassava trasts with formulations CCSF2-CCSF5, which is associated with mouth-
and increased colour intensity (khaki). Likewise, Tegeye, Kaur, Kaur, feel, colour, and oiliness attributes. Conversely, formulations CF and
and Singh (2019) reported that composite flours prepared from differ- CCSF1 correlated with high loadings of aroma, jellying, and sweetness
ent natural nutrient-dense food products improve consumers’ liking of attributes. PC 2 contrasts with the formulation CCSF1- CCSF3, associ-
colour. Moreover, the presence of a high amount of fats in chia seeds ated with high aroma, colour, and oiliness, and the other side sample
of 30 - 35% (Suri, Passi, & Goyat, 2016) increased oiliness, colour, and CCSF5 and CF are associated with high loading of jellying, mouthfeel,
overall acceptability intensities in composite porridge samples (CCSF3- sweetness, and viscosity. These results show that the distribution of com-
CCSF5) relative to the whole cassava porridge. These results show that posite porridges in the PCA Bi-plot is linked to attributes influenced by
incorporating chia seeds flour might have induced the overall accept- the percentage of chia seeds flour incorporated in the blend.
ability of the composite samples witnessed in the study. These results
concur with the findings by Emmaculate et al. (2020), Eze, Okafor, 3.4.2. Relationship between descriptive data and hedonic liking by PLSR by
Omah and Azuka (2020), Ramírez-Jiménez, Gaytán-Martínez, Morales- external preference mapping
Sánchez and Loarca-Piña (2018). Contrary to our expectations, a high The relationship between descriptive data (mean intensity scores)
preference for porridge samples at higher substitution levels of cassava and hedonic liking (customers acceptability) is presented by partial least
flour augmented the overall acceptability registered in the current study. square regression (PLSR) (Fig. 3). The PLSR showed that many con-
In addition, the results affirm the appeal of consumers increased after sumers fall to the right of the vertical Y-axis. The liking of the panel-
they had tasted these samples. lists goes in the direction of preference of porridge samples of CCSF5,

5
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

Fig. 2. Bi-plot from PCA of descriptive sensory data for porridge samples.

Fig. 3. Correlation loadings from a Partial

Least Square Regression of porridge samples
with descriptive data as X variables and hedo-
nic rating as Y variables.

CCSF4, and CCSF3, followed by CCSF2. These findings suggest that the the variability of descriptive sensory data (Fig. 2) of porridge samples
main three attributes, mouthfeel, colour, and oiliness, were the main among the consumers.
drivers of cassava-chia seeds porridge samples by the study’s panellist. The Partial Least Square Regression’s preference mapping also re-
On the other hand, the high association between the CF sample and high vealed findings associated with mouthfeel, viscosity, oiliness, and colour
jellying, sweetness, and dried cassava aroma attributes could account for as drivers for liking the porridges with high cassava flour substitution
the low consumer acceptability. with chia seeds flour, i.e., CCSF3, CCSF4, and CCSF5. Oiliness emerged
Furthermore, the effect of each attribute on porridge liking is pre- to contribute positively but weakly compared to colour and mouthfeel
sented by the PLSR plot (Fig. 4). Mouthfeel, colour, and oiliness at- to the product’s overall liking due to the consumer’s perception of food
tributes contributed strongly and positively to the sample acceptability, products with a high amount of fat, as observed by (Ballco, Jurado, &
which falls to the vertical Y-axis’ right, supporting the results in Fig. 4. Gracia, 2020).
However, the viscosity attribute also contributed positively, albeit not Appearance, colour, and taste constitute an index of the inherent
as strong as other attributes (mouthfeel, colour, and oiliness). On the good quality of foods and significantly influence food acceptability,
contrary, dried cassava aroma, sweetness, and jellying attributes con- selection, consumption, ingestion, and satiation (Barrett, Beaulieu, &
tributed strongly and negatively to the acceptance of porridge samples. Shewfelt, 2010; Dias et al., 2012; Mongi, Bernadette, Chove, & Wick-
These results, which affirmed a relationship between sensory attributes, lund, 2013). The negative association observed between whole cassava
samples, and panellists (Fig. 4), support the results of customer accept- porridge (control), and its associated flavour (aroma and sweetness) sig-
ability (Table 5), and customer’s preference (Fig. 1), and directions of nals the unappealing characteristics of these attributes to consumers.

6
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

Fig. 4. PLSR plot shows the effect of each sensory attribute on consumer liking for porridge samples.

Thus, the product development of cassava-chia seeds composite por- Alamu, E. O., Maziya-Dixon, B., & Dixon, A. G. (2017). Evaluation of proximate composi-
ridges showed appealing sensory attributes, especially colour, mouth- tion and pasting properties of high-quality cassava flour (HQCF) from cassava geno-
types (Manihot esculenta Crantz) of 𝛽-carotene-enriched roots. LWT, 86, 501–506.
feel, and oiliness. More significant considerations of consumer accept- 10.1016/j.lwt.2017.08.040.
ability, marketability, and consumption of the product appear essential. Alphonce, S., Kaale, L. D., Millinga, F., & Rweyemamu, L. M. (2020). Enrichment of fer-
mented cassava meal ‘mchuchume’ with micronutrient ingredients from Soya bean
flour and Moringa oleifera leaves powder. Journal of the Science of Food and Agricul-
4. Conclusion ture. 10.1002/jsfa.10984.
Alyami, J., Ladd, N., Pritchard, S. E., Hoad, C. L., Sultan, A. A., Spiller, R. C., & Mar-
ciani, L. (2019). Glycaemic, gastrointestinal and appetite responses to breakfast por-
Cassava flour used to make porridge is generally associated with low
ridges from ancient cereal grains: A MRI pilot study in healthy humans. Food research
nutritional composition due to high starch content. Formulation of com- international, 118, 49–57. 10.1016/j.foodres.2017.11.071.
posite flour of cassava mixed with chia seeds flour, on the other hand, Amerine, M. A., Pangborn, R. M., & Roessler, E. B. (2013). Principles of sensory evaluation
of food. Elsevier.
enhances the sensory attributes compared to non-blended cassava flour
Aso, S. N., Teixeira, A. A., & Welt, B. A. (2015). Physical properties of cassava flour made
itself. Thus, blending 15-25% of chia seeds flour with cassava flour im- from solar convection-dried cassava chips. Applied Engineering in Agriculture, 31(4),
proved intensities of colour and mouthfeel and reduced sweetness and 655–660. 10.13031/aea.31.10573.
aroma associated with dried cassava flour. The results also show that Ballco, P., Jurado, F., & Gracia, A. (2020). Do health claims add value to nutritional
claims? Evidence from a close-to-real experiment on breakfast biscuits. Food Qual-
mouthfeel, viscosity, oiliness, and colour were the drivers behind the ity and Preference, Article 103968. 10.1016/j.foodqual.2020.103968.
consumers’ liking and porridge preference. This study, therefore, calls Barrett, D. M., Beaulieu, J. C., & Shewfelt, R. (2010). Color, flavor, texture, and nutritional
for the formulation of composite flour of cassava blended with chia seeds quality of fresh-cut fruits and vegetables: desirable levels, instrumental and sensory
measurement, and the effects of processing. Critical reviews in food science and nutrition,
(15-25%) to boost the sensory attributes and nutritional composition 50(5), 369–389. 10.1080/10408391003626322.
due to the richness of nutrients found in chia seeds. This study also cre- Batey, I. L., & Curtin, B. M. (2000). Effects on pasting viscosity of starch and flour from
ates an opportunity for small to large-scale food processors to valorise different operating conditions for the Rapid Visco Analyser. Cereal Chemistry, 77(6),
754–760. 10.1094/CCHEM.2000.77.6.754.
cassava flour by producing cassava-chia seeds composite flour benefi- Bayata, A. (2019). Review on nutritional value of cassava for use as a staple food. Science
cial to the broader population in terms of food security and nutritional Journal of Analytical Chemistry, 7(4), 83–91. 10.11648/j.sjac.20190704.12.
enhancement. Civille, G. V., & Carr, B. T. (2015). Sensory evaluation techniques. CRC Press.
10.1201/b19493.
Coelho, M. S., & Salas-Mellado, M. d. l. M. (2014). Chemical characterization of chia
Declaration of Competing Interest (Salvia hispanica L.) for use in food products. Journal of Food and Nutrition Research,
2(5), 263–269. 10.12691/jfnr-2-5-9.
Dias, N. A. A., Lara, S. B., Miranda, L. S., Pires, I. S. C., Pires, C. V., & Halboth, N. V. (2012).
The authors declare that they have no conflict of interest. Influence of color on acceptance and identification of flavor of foods by adults. Food
Science and Technology, 32(2), 296–301.
Emmaculate, S., Wandayi, O. M., Ooko, A. G., & Kadenyeka, M. V. (2020). Consumers
Acknowledgments acceptability of extruded maize-sorghum composite flours fortified with grain ama-
ranth, baobab, and orange-fleshed sweet potatoes. African Journal of Food Science,
The authors are grateful to the Tanzania Bureau of Standards (TBS) 14(9), 274–284. 10.5897/AJFS2020.1996.
Eze, C., Okafor, G., Omah, E., & Azuka, C. (2020). Micronutrients, antinutrients composi-
for its financial support to her employee for fulfilling the requirements tion, and sensory properties of extruded snacks made from sorghum and charamenya
for the degree of Doctor of Philosophy (Food Science and Technology) flour blends. African Journal of Food Science, 14(1), 25–31.
of the University of Dar es Salaam. Grosshagauer, S., Milani, P., Kraemer, K., Mukabutera, A., Burkon, A., Pignitter, M., &
Somoza, V. (2020). Inadequacy of nutrients and contaminants found in porridge-
type complementary foods in Rwanda. Maternal & Child Nutrition, 16(1), e12856.
References 10.1111/mcn.12856.
Imakumbili, M. L., Semu, E., Semoka, J. M., Abass, A., & Mkamilo, G. (2019). Farmers’
Ajifolokun, O., Basson, A., Osunsanmi, F., & Zharare, G. (2019). Nutritional composition perceptions on the causes of cassava root bitterness: A case of konzo-affected Mtwara
and organoleptic properties of composite maize porridge. Journal of Food Processing region, Tanzania. PloS one, 14(4), Article e0215527. 10.1371/journal.pone.0215527.
Technology, 10(798), 2.

7
A. Katunzi-Kilewela, R.J. Mongi Dr, L.D. Kaale Dr et al. Applied Food Research 2 (2022) 100038

Ixtaina, V. Y., Nolasco, S. M., & Tomas, M. C. (2008). Physical properties of Pomeranz, Y. (2012). Functional properties of food components. Academic Press.
chia (Salvia hispanica L.) seeds. Industrial crops and products, 28(3), 286–293. Ramírez-Jiménez, A. K., Gaytán-Martínez, M., Morales-Sánchez, E., & Loarca-
10.1016/j.indcrop.2008.03.009. Piña, G. (2018). Functional properties and sensory value of snack bars added
Laswai, H., Silayo, V., Mpagalile, J., Balegu, W., & John, J. (2006). Improvement and pop- with common bean flour as a source of bioactive compounds. LWT, 89, 674–680.
ularization of diversified cassava products for income-generating and food security: a 10.1016/j.lwt.2017.11.043.
case study of Kibabu. African Journal of Food, Agriculture, Nutrition and Development, Richter, M., Baerlocher, K., Bauer, J. M., Elmadfa, I., Heseker, H., Leschik-Bonnet, E., &
6(1), 1–15. 10.4314/ajfand.v6i1.19168. Stehle, P. (2019). Revised reference values for the intake of protein. Annals of Nutrition
Lawless, H. T., & Heymann, H. (2010a). Principles of good practice Sensory Evaluation of and Metabolism, 74(3), 242–250. 10.1159/000499374.
Food (pp. 57–77). Springer. 10.1007/978-1-4419-6488-5_3. Segura-Campos, M. R., Ciau-Solís, N., Rosado-Rubio, G., Chel-Guerrero, L., & Betancur-
Lawless, H. T., & Heymann, H. (2010b). Sensory evaluation of food: principles and practices. Ancona, D. (2014). Chemical and functional properties of chia seed (Salvia hispanica
Springer Science & Business Media. 10.1007/978-1-4419-6488-5. L.) gum. International journal of food science 2014. 10.1155/2014/241053.
Lawless, H. T., & Heymann, H. (2013). Sensory evaluation of food: principles and practices. Suri, S., Passi, S. J., & Goyat, J. (2016). Chia seed (Salvia hispanica L.)—A new age func-
Springer Science & Business Media. tional food. Paper presented at the 4th International Conference on Recent Innovations in
Marcinek, K., & Krejpcio, Z. (2017). Chia seeds (Salvia hispanica): health promoting prop- Science Engineering and Management.
erties and therapeutic applications-a review. Roczniki Państwowego Zakładu Higieny, Tappiban, P., Ying, Y., Pang, Y., Sraphet, S., Srisawad, N., Smith, D. R., & Bao, J. (2020).
68(2). Gelatinization, pasting, and retrogradation properties and molecular fine structure
Martens, H., & Martens, M. (2001). Multivariate analysis of quality. An introduction. IOP of starches from seven cassava cultivars. International Journal of Biological Macro-
Publishing. 10.1088/0957-0233/12/10/708. molecules, 150, 831–838. 10.1016/j.ijbiomac.2020.02.119.
Mongi, R. J., Bernadette, N., Chove, B., & Wicklund, T. (2013). Descriptive sensory anal- Taylor, J., & Emmambux, M. (2008). Products containing other speciality grains: sorghum,
ysis, consumer liking, and preference mapping for solar dried mango cv Dodo. Food the millets and pseudocereals Technology of functional cereal products (pp. 281–335).
Sci. Qual. Manag, 16, 16–23. Elsevier. 10.1533/9781845693886.2.281.
Mtunguja, M. K., Beckles, D. M., Laswai, H. S., Ndunguru, J. C., & Sinha, N. J. (2019). Tegeye, M., Kaur, A., Kaur, J., & Singh, H. (2019). Value-added convenience food from
Opportunities to commercialize cassava production for poverty alleviation and im- composite sorghum-maize-sweet potato flour blends. Indian Journal of Agricultural Sci-
proved food security in Tanzania. African Journal of Food, Agriculture, Nutrition and ences, 89(11), 1906–1910.
Development, 19(1), 13928–13946. Tomic, O., Luciano, G., Nilsen, A., Hyldig, G., Lorensen, K., & Næs, T. (2010). Analysing
Muñoz, L., Cobos, A., Diaz, O., & Aguilera, J. (2012). Chia seeds: microstructure, sensory panel performance in a proficiency test using the PanelCheck software. Euro-
mucilage extraction, and hydration. Journal of Food Engineering, 108(1), 216–224. pean Food Research and Technology, 230(3), 497–511. 10.1007/s00217-009-1185-y.
10.1016/j.jfoodeng.2011.06.037. Ubeda, C., Callejón, R., Troncoso, A., & Morales, M. (2017). Consumer acceptance of
Muoki, P. N. (2013). Nutritional, rheological, and sensory properties of extruded cassava-soy new strawberry vinegar by preference mapping. International journal of food properties,
complementary porridges. The University of Pretoria. 20(11), 2760–2771. 10.1080/10942912.2016.1252388.
Nout, M. J. R., & Rombouts, F. M. (1992). Fermentative preservation of plant foods. J. Yenket, R. (2011). Understanding methods for internal and external preference mapping and
Appl. Bacteriol, 73, 136S–147S. clustering in sensory analysis. Kansas State University.
Onyango, C., Luvitaa, S. K., Unbehend, G., & Haase, N. (2020). Nutrient composi- Zhu, F., & Chan, C. (2018). Effect of chia seed on glycemic response, texture, and sensory
tion, sensory attributes, and starch digestibility of cassava porridge modified with properties of Chinese steamed bread. LWT, 98, 77–84. 10.1016/j.lwt.2018.08.016.
hydrothermally-treated finger millet. Journal of Agriculture and Food Research, 2, Ar-
ticle 100021. 10.1016/j.jafr.2020.100021.
Otondi, E. A., Nduko, J. M., & Omwamba, M. (2020). Physico-chemical properties of ex-
truded cassava-chia seed instant flour. Journal of Agriculture and Food Research, 2,
Article 100058. 10.1016/j.jafr.2020.100058.