Journal of Human Nutrition and Dietetics
PUBLIC HEALTH NUTRITION AND EPIDEMIOLOGY Consumer behaviour towards vegetables: a study on domestic processing of broccoli and carrots by Dutch households R. Bongoni, R. Verkerk, M. Dekker & L. P. A. Steenbekkers Food Quality and Design, Wageningen University, Wageningen, the Netherlands
Keywords cluster analysis, domestic processing, health, selfreporting questionnaire, taste, texture. Correspondence R. Bongoni, Food Quality and Design, Wageningen University, Bornse Weilanden 9, 6708WG, the Netherlands. Tel.: +31 317 48 25 20 E-mail: [email protected] How to cite this article Bongoni R., Verkerk R., Dekker M. & Steenbekkers B. (2014) Consumer behaviour towards vegetables: a study on domestic processing of broccoli and carrots by Dutch households. J Hum Nutr Diet. 28, 219–225 doi: 10.1111/jhn.12245
Abstract Background: Preferences for sensory properties (e.g. taste and texture) are assumed to control cooking behaviour with respect to vegetables. Conditions such as the cooking method, amount of water used and the time-temperature profile determine the nutritional quality (e.g. vitamins and phytochemicals) of cooked vegetables. Information on domestic processing and any underlying motives can be used to inform consumers about cooking vegetables that are equally liked and are nutrient-rich. Methods: Two online self-reporting questionnaires were used to identify domestic processing conditions of broccoli and carrots by Dutch households. Questions on various aspects of domestic processing and consumer motives were included. Descriptive data analysis and hierarchical cluster analysis were performed for both vegetables, separately, to group consumers with similar motives and behaviour towards vegetables. Results: Approximately 70% of consumers boiled vegetables, 8–9% steamed vegetables, 10–15% stir fried raw vegetables and 8–10% stir fried boiled vegetables. Mainly texture was used as a way to decide the ‘doneness’ of the vegetables. For both vegetables, three clusters of consumers were identified: texture-orientated, health-orientated, or taste-orientated. The texture-orientated consumers are identified as the most prevalent (56–59%) group in the present study. Statistically significant associations are found between domestic processing conditions and clusters, whereas no such association are found between demographic details and clusters. Conclusions: A wide variation in domestic processing of broccoli and carrots is found in the present study. Mainly sensory properties (i.e. texture and taste) determined the domestic processing conditions. The findings of the present study can be used to optimise cooking to yield vegetables that meet consumer’s specific sensory preference and are higher in nutrients, and as well as to communicate with target consumer groups.
Introduction Vegetables are processed at the domestic level before consumption. Domestic processing brings about a significant change in the sensorial properties and nutritional quality of vegetables (Kala & Prakash, 2006). Depending on the conditions applied, nutritional quality decreases, ª 2014 The British Dietetic Association Ltd.
e.g. amount of nutrients and phytochemicals, although the bioavailability of certain compounds such as carotenes increases (Dekker et al., 2000). Meanwhile, changes in the sensorial properties such as texture and colour are assumed to be controlled by consumers to meet their (and their family’s) sensory preferences. Therefore, it can be hypothesised that motives such as sensory preferences 219
R. Bongoni et al.
Consumer behaviour towards vegetables
towards vegetables and the habits of consumers determine the cooking conditions (Grunert, 1995; Moskowitz, 1995). When such motives and consumer behaviour towards vegetables are known, food technologists can assess optimised cooking that will result in vegetables (products) that are liked equally but at the same time are higher in nutritional quality (Bongoni et al., 2013). However, for process optimisation detailed information on consumer behaviour towards vegetables and the underlying motives must be known. To our knowledge, there is a lack of such information in the literature. In addition to process optimisation, such information can also be used to cluster consumers into homogenous groups comparable in behaviour and motives (Blanchette & Brug, 2005; Demydas, 2011; Simunaniemi et al., 2013; Wang et al., 2013). Two separate online self-reporting questionnaires are used to identify domestic processing conditions for broccoli and carrots among Dutch households. Broccoli and carrots are chosen for the present study because they are among the commonly consumed vegetables in the Netherlands (Voorrips et al., 2000) and differ morphologically from each other. Both vegetables are rich in a variety of phytochemicals such as polyphenols, glucosinolates and vitamin C in broccoli (Verkerk et al., 2009) as well as phytochemicals such as b-carotene, falcarinol and polyphenols in carrots (Griep et al., 2011; Seljasen et al., 2013). Such compounds are reported to exhibit anti-oxidant and anticarcinogenic activity (Voorrips et al., 2000; Verkerk et al., 2009). In the first part of this paper domestic processing of broccoli and carrots applied by consumers is reported. Next, the results of cluster analysis identifying groups of consumers with comparable domestic processing conditions and motives are shown for both vegetables separately. Materials and methods Online self-reporting questionnaire Two separate online self-reporting questionnaires, which were adapted from the validated questionnaire described by Bongoni et al. (2014), were used to identify domestic processing conditions of broccoli and carrots among Dutch households. Thirteen closed-ended questions were used on the cooking method, part/size of the vegetable piece, water level and type of start used when boiling, cooking time, addition of salt and ways to decide the ‘doneness’ of the vegetable. Water level in the pan for boiling is categorised as: low or high water level (see Supporting Information, Fig. S1). The type of start in the study is categorised as: cold or hot start. In a cold start cooking time commences as the vegetables are heated along with the tap water (approximately 22 °C) in a pan. Cooking time in a hot start commences with the addition of the vegetables into the boiling water. Questions related to the motives for some practices 220
(e.g. water level, type of start, addition of salt) were included. Demographic details (e.g. age, gender, family composition, education) were asked at the end of the questionnaire. The web links to the online surveys were spread among the personal contacts and through the database of Food and Biobased Research (FBR), Wageningen UR, the Netherlands to obtain a large number and maximum variation (in terms of geographical location) of respondents. Participants The target group comprised the main meal preparer of Dutch households. In total, the data from 567 consumers on broccoli and the data from 348 consumers on carrot domestic processing conditions and underlying motives were used for the present study. In the case of broccoli, 12 respondents microwaved and three respondents pressure-cooked the vegetables. In the case of carrots, three respondents microwaved, two pressure-cooked the vegetables and 74 ate carrots raw (and consequently answered all the other questions as ‘none of the above’ or ‘notapplicable’). Given the frequency of these cooking methods and given that the questionnaire asked for domestic processing conditions of vegetables, the data from these participants along with the data from incomplete questionnaires were excluded, resulting in 567 respondents for broccoli and 348 for carrots. The characteristics of the study sample are shown in Table 1. Statistical analysis Statistical analysis was performed using IBM SPSS STATISTICS, version 21(IBM Corp., Armonk, NY, USA). Descriptive statistical analysis was performed to summarise the domestic processing of broccoli and carrots. Cluster analysis was performed to identify consumer groups with similar Table 1 Characteristics of the study sample n (%) Demographic variable Gender Male Female Household type Students (20–25 years) One-person household Couple with children Couple without children Elderly (>65 years) Education level Secondary vocational Higher professional University
Broccoli (n = 567)
Carrots (n = 348)
198 (35) 369 (65)
100 (29) 248 (71)
62 60 208 116 121
38 27 131 96 56
(11) (11) (35) (22) (21)
132 (23) 215 (38) 220 (39)
(11) (8) (37) (28) (16)
96 (28) 132 (38) 120 (34)
ª 2014 The British Dietetic Association Ltd.
R. Bongoni et al.
characteristics (Demydas, 2011) (i.e. domestic processing conditions). Hierarchical cluster analysis was performed with Ward’s method at squared Euclidean distance (Simunaniemi et al., 2013) on variables: part of the vegetable piece, cooking method, water level and, type of start for boiling method and cooking time. The number of clusters was decided based on the dendrogram and logical interpretations of the solutions. There was no substantial correlation (P < 0.05) between the input variables. The clusters were described based on domestic vegetable processing conditions and motives. A chi-squared test was performed to determine statistically significant associations between the clusters and domestic processing variables (P < 0.05).
Consumer behaviour towards vegetables Table 2 Summary of the domestic vegetables applied by Dutch households Aspects on domestic processing of vegetables
Clustering consumers based on domestic processing of broccoli Cluster 1: consumers who are ‘hard texture’-orientated This cluster comprises 39% (n = 221) of the consumers from the study sample who are orientated towards having a ‘hard texture’ of the broccoli. ª 2014 The British Dietetic Association Ltd.
of
Details for carrots (%)
Cooking method
Boiling (71) Steaming (9) Stir frying (10) Boiling followed by stir frying (10)
Boiling (69) Steaming (8) Stir frying (15) Boiling followed by stir frying (8)
Part of the vegetable used
Florets without stalks (10)
Whole carrots (61)
Florets with stalks (72) Florets with stalks and sliced stem (18%)
Cut into pieces (39)
Low water level (18)
Low water level (17)
High water level (82)
High water level (83)
Cold (60)
Cold (73)
Hot (40)
Hot (27)
Cooking time
1–5 min (19) 6–10 min (49) 11–15 min (20) 16–20 min (8) >20 min (4)
1–5 min (7) 6–10 min (40) 11–15 min (27) 16–20 min (21) >20 min (5)
Ways to decide doneness
Texture (64)
Texture (65)
Colour (6) Taste (19) Time (11)
Colour (2) Taste (25) Time (8)
Yes (27)
Yes (29)
Water level when vegetables were boiled
Type of start when boiling
Results Domestic processing of vegetables by consumers Table 2 presents the domestic vegetable processing conditions applied by Dutch households. As expected, a wide variation in the domestic processing of vegetables is observed (Table 2). Approximately 70% of consumers in the present study cooked broccoli and carrots mainly by boiling. Stir frying (10% for broccoli and 15% for carrots), boiling followed by stir frying (10% for broccoli and 8% for carrots) and steaming (9% for broccoli and 8% for carrots) were the other cooking methods used for broccoli and carrots. Most (>80%) consumers reported using a high water level to boil their vegetables. A cold start was applied by 60% and 72% of consumers for boiling broccoli and carrots, respectively. Texture was used by approximately 65% of the consumers to determine the ‘doneness’ of the vegetable during cooking. Except for the type of start, no substantial differences were found in domestic processing between the two vegetables.
conditions
Details for broccoli (%)
Ethical approval The research used online self-reporting questionnaires for information on the domestic processing of vegetables. The respondents were free to participate. Because the research is non-invasive and the details of the participants remain undisclosed, this research does not fall within the remit of the Helsinki Declaration. However, this research received ethical approval from the Social Sciences Ethics Committee, Wageningen University, the Netherlands and the project complies with the Netherlands Code of Conduct for Scientific Practice.
processing
Addition of salt
These consumers boiled the broccoli (100%) using a high water level (100%) and either a cold start (53%) or a hot start (47%). The water level and the type of start used were motivated as a habitual practice. Broccoli is cooked for 6–10 min (77%). Texture change during cooking was used to decide on the ‘doneness’ of broccoli (65%) and 28% of consumers from this group added salt. Cluster 2: consumers who are health-orientated This cluster comprises 33% (n = 188) of consumers from the study sample who are orientated towards having higher health benefits as a result of the consumption of broccoli. This cluster constitutes consumers who use different cooking methods: boiling (40%), steaming (27%), stir 221
Consumer behaviour towards vegetables
frying (29%) and boiling followed by stir frying (4%). When methods such as boiling or boiling followed by stir frying were used, a low water level was used (100%) to prevent nutrient losses into the cooking water. A cold start (100%) was used as a habitual practice. Short to medium cooking times: 1–5 min (26%) and 6–10 min (46%) were used by this group of consumers. Again, the texture of broccoli during cooking was used to decide on ‘doneness’ (56%). Compared to the total study sample (Table 2), a lower number of consumers (16%) added salt to the broccoli. Motives for not adding salt were to restrict the salt intake and to perceive the vegetable taste. Cluster 3: consumers who are ‘medium to soft texture’-orientated This cluster comprises 20% (n = 114) of consumers from the study sample who are orientated towards having a ‘medium to soft’ texture of the broccoli. These consumers mostly boiled the broccoli (97%) using a high water level (100%) and a cold start (70%). Again, habit was the reason for applying this water level and the type of start. Broccoli was cooked for 11–15 min (57%) or for 16–20 min (28%), which can result in a ‘medium to soft’ texture of broccoli. Texture of broccoli during cooking was used to decide on ‘doneness’ (66%). In this cluster, 39% of consumers added salt, which is a higher number than in the total sample (Table 2). Cluster 4: consumers who are taste-orientated This cluster comprises 8% (n = 44) of consumers from the study sample who are orientated towards the taste of the broccoli. These consumers used boiling followed by stir frying (100%) to cook the broccoli. When boiling, a high water level was used (100%) as a habit and mainly a hot start was used (69%) to speed-up the cooking process. Cooking times of 1–5 min (30%) or 6–10 min (52%) were used. Besides texture (55%), taste (40%) was used to assess the ‘doneness’ of the broccoli, which is different from the total sample (Table 2). In this cluster, 39% of the consumers used salt for taste reasons, which is a higher number compared to the total sample (Table 2). Clustering consumers based on domestic processing of carrots Cluster 1: consumers who are texture-orientated This cluster comprises 56% (n = 196) of consumers from the study sample who are orientated towards the texture of the cooked carrots. Carrots are cooked by boiling (100%) using a high water level (100%) and a cold start (100%) as a habitual pr222
R. Bongoni et al.
actice. Carrots are cooked over a range of time that can result in ‘hard to soft’ carrots. In this cluster, 35% of the consumers cooked carrots for 6–10 min, 29% for 11– 15 min and 29% for 16–20 min. Texture of carrots was used to decide on ‘doneness’ by 71% of the consumers. Salt was added habitually by 38% of the consumers, which is higher than in the total sample (Table 2). Cluster 2: consumers who are health-orientated This cluster comprises 18% (n = 63) of consumers from the study sample who are orientated towards the health benefits of the consumption of carrots. Carrots were cooked either by boiling (65%), steaming (16%) or stir frying (19%). To boil, a low water level was used to prevent nutrient losses and a cold start was used as a habitual practice. Carrots were cooked for 6–10 min by 28% of the consumers and for 11–15 min by 38% of the consumers. Texture of the carrots was used (67%) to decide on ‘doneness’. A relatively low number of consumers (13%) added salt in comparison to the total sample (Table 2). Motives for not adding salt were to restrict salt intake and to perceive the carrot taste. Cluster 3: consumers who are taste-orientated and want ‘hard texture’ carrots This cluster comprises 18% (n = 63) of consumers from the study sample who are orientated towards the taste of carrots that are ‘hard’ in texture. Consumers of this cluster used different methods to cook carrots: stir frying (69%), steaming (26%) and boiling followed by stir frying (5%). A cooking time of 6– 10 min was used by most of the consumers (69%). Texture (42%) or taste (40%) was used to determine the ‘doneness’ of carrots. This cluster has a higher percentage of salt-users (37%) compared to the total sample (Table 2). Cluster 4: consumers who are taste-orientated and want ‘medium to soft texture’ carrots This cluster comprises 8% (n = 26) of consumers from the study sample who are orientated towards the taste of carrots that are ‘medium to soft’ in texture. These consumers used boiling followed by stir frying (100%) to cook carrots. For boiling carrots, these consumers used a high water level (100%) and either a cold start (53%) or a hot start (47%) as a habitual practice. The carrots were first boiled for 6–10 min (50%) or for 11–15 min (27%), which was followed by stir frying. Texture (46%) or taste (54%) of the carrots was used to decide on ‘doneness’. This cluster has the highest percentage of salt-users (39%) in comparison to the total sample (Table 2).
ª 2014 The British Dietetic Association Ltd.
R. Bongoni et al.
Association between the input variables of the cluster analysis and the clusters The input variables of the cluster analysis (i.e. the domestic processing conditions) and the clusters from the cluster analysis showed a statistically significant association with each other, indicating that these variables and the clusters are related (see Supporting Information, Table S1). While for both vegetables, the demographic details of the respondents did not show a statistically significant association with the clusters (see Supporting Information, Table S2).
Discussion To our knowledge, the present study is the first to investigate consumer behaviour towards vegetables (i.e. processing of vegetables at the domestic level) and to cluster consumers based on their reported behaviour and motives. Most (70%) consumers cooked vegetables by boiling. Boiling compared to steaming results in a higher loss of (water-soluble) phytochemicals by leaching into the cooking water (Kala & Prakash, 2006). Leaching depends on the vegetable-water ratio used and the cooking timetemperature profile (Verkerk et al., 2009). Although steaming is reported to reduce the loss of phytochemicals (Bernhardt & Schlich, 2006; Yuan et al., 2009), only a few consumers (8–9%) in the present study steamed vegetables. Stir frying, where the temperature of cooking reaches approximately to 140 °C, is detrimental to a great extent for phytochemical stability (Yuan et al., 2009). Approximately 10–15% of the consumers in the present study stir fried vegetables. Texture, which is reported to determine the acceptability of foods (Poelman & Delahunty, 2011), is also used as the main indicator to determine the ‘doneness’ of the vegetables in the present study. Weegels & van Veen (2001) reported a wide variation in behaviour between consumers when performing the same household activity. Such wide variation in behaviour between consumers is also observed when processing vegetables at the domestic level. The quality of cooked vegetables at the time of consumption depends on the way that they were processed (Kala & Prakash, 2006) (e.g. at the domestic level). The information obtained on variation in the domestic processing of vegetables will help to identify aspects of domestic processing that might determine the health and sensory properties of cooked vegetables. In the present study, cluster analysis presented three groups of consumers for both vegetables: those who are texture-orientated, health-orientated or taste-orientated. In the case of broccoli, the texture-orientated consumers and ª 2014 The British Dietetic Association Ltd.
Consumer behaviour towards vegetables
in the case of carrots, the taste-orientated consumers were divided further based on the applied cooking time that results in different textures. The texture-orientated consumers (clusters 1 and 3 in the case of broccoli and cluster 1 in the case of carrots) are the most prevalent group comprising 56–59% of the study sample. These clusters of consumers mostly used a high water level with a cold start and often cooked the vegetables longer than the total sample does (i.e. 6– 10 min) (Table 2), except for the ‘hard texture’-orientated consumers of broccoli. Additionally, compared to the total sample, this cluster of consumers consists of a relatively high percentage of salt-users (28–39%). The behaviour of these consumers comprises mostly habitual practices, which could arise as a result of performing a behaviour repeatedly over time (Simunaniemi et al., 2013). Furst et al. (1996) reported that a past behaviour or experience greatly influences the current behaviour of the consumer. The quality of a product that a consumer experienced or gained (Furst et al., 1996) (in this case, a specific texture of a vegetable that was obtained in previous cooking sessions) might have led to a repeated behaviour in the next cooking sessions. This repeated behaviour probably turned into a habit, for which consumers find it difficult to explain details and specific factors that contribute to their behaviour (Brug et al., 1995; Simunaniemi et al., 2013). Based on the prevalence of this group and the processing conditions that can be detrimental to (nutrients and) phytochemicals, these consumers can be assisted with information on optimised processing conditions, which will result in vegetables within the range of an ‘acceptable texture’ but at the same time are higher in nutritional quality. The health-orientated consumers counted for 20–30% of the study sample for both vegetables. Steaming was used by a higher number of consumers in this group: 27% for broccoli and 16% for carrots, compared to the total sample (Table 2). When applicable, a low water level was used to prevent nutrient losses during cooking and shorter cooking times were applied for broccoli and slightly longer for carrots (with respect to the total sample as shown in Table 2), given the morphological difference between vegetables. Fewer consumers used salt to restrict their salt intake. Grunert (2002) and Brunsø et al. (2002) reported that consumers relate healthiness to eating healthy foods and to avoiding unhealthy foods, aiming to fulfil their basic life values such as good health, a long life and family welfare (Geeroms et al., 2008). The taste-orientated consumers (cluster 4 for broccoli and clusters 3 and 4 for carrots) mainly stir fried the raw or boiled vegetables, cooked with a high water level and mainly applied a hot start, all of which can result in high nutritional quality losses of vegetables. Taste was used to 223
Consumer behaviour towards vegetables
assess the ‘doneness’ of vegetables and was also the reason for the use of salt. The limitation of the present study is that the online self-reporting questionnaires used are not checked for reliability and validity, although they are adapted from the reliable and valid self-reporting questionnaire described by Bongoni et al.(2014). Any adapted questionnaire should be validated prior to administration (Barton et al., 2011). In summary, the findings of the present study are novel and can be used (by food technologists) to optimise cooking, which results in vegetables that meet consumers’ sensory preferences (e.g. texture) and higher in nutrients (Bongoni et al., 2013). Such information on optimised cooking developed for every consumer cluster identified in the present study will assist these consumers to adapt their existing cooking practices. This cluster-specific information can be communicated to consumers in different ways (e.g. through forums, blogs, recipe books or television programmes or cooking classes and by placing pamphlets in the vegetable shops or supermarkets). Cooking programmes at schools could be organised that shall assist in inculcating ‘healthy cooking habits’ in children. To generalise the consumer clusters found in the present study, further research is recommended to cluster and compare consumer behaviour towards vegetables of the same morphology (e.g. potatoes to carrots or cauliflower to broccoli) and of different morphology (e.g. green leafy vegetables). Acknowledgments The present study was supported financially by the VLAG graduate school, Wageningen University, the Netherlands.
Conflict of interests, source of funding and authorship The authors declare that they have no conflict of interests. No funding declared. RB collected the data, analysed, interpreted and wrote the manuscript. All the authors together designed the study, interpreted the data and contributed to writing and critically reviewing the manuscript submitted for publication.
References Barton, K.L., Wrieden, W.L. & Anderson, A.S. (2011) Validity and reliability of a short questionnaire for assessing the
224
R. Bongoni et al.
impact of cooking skills interventions. J. Hum. Nutr. Diet. 24, 588–595. Bernhardt, S. & Schlich, E. (2006) Impact of different cooking methods on food quality: retention of lipophilic vitamins in fresh and frozen vegetables. J. Food Eng. 77, 327–333. Blanchette, L. & Brug, J. (2005) Determinants of fruit and vegetable consumption among 6–12-year-old children and effective interventions to increase consumption. J. Hum. Nutr. Diet. 18, 431–443. Bongoni, R., Steenbekkers, L.P.A., Verkerk, R., van Boekel, M.A.J.S. & Dekker, M. (2013) Studying consumer behaviour related to the quality of food: a case on vegetable preparation affecting sensory and health attributes. Trends Food Sci. Technol. 33, 139–145. Bongoni, R., Verkerk, R., Dekker, M. & Steenbekkers, L.P.A. (2014) Evaluation of research methods to study domestic food preparation. Br. Food J. [Epub ahead of print]. Brug, J., Lechner, L. & De Vries, H. (1995) Psychosocial determinants of fruit and vegetable consumption. Appetite 25, 285–296. Brunsø, K., Fjord, T.A. & Grunert, K.G. (2002) Consumer’s food choice and quality perception, working paper no. 77, The MAPP Centre: The Aarhus School of Business. Dekker, M., Verkerk, R. & Jongen, W.M.F. (2000) Predictive modeling of health aspects in the food production chain: a case study on glucosinolates in cabbage. Trends Food Sci. Technol. 11, 174–181. Demydas, T. (2011) Consumer segmentation based on the level and structure of fruit and vegetable intake: an empirical evidence for US adults from the National Health and Nutrition Examination Survey (NHANES) 2005–2006. Public Health Nutr. 14, 1088–1095. Furst, T., Connors, M., Bisogni, C.A., Sobal, J. & Falk, L.W. (1996) Food choice: a conceptual model of the process. Appetite 26, 247–266. Geeroms, N., Verbeke, W. & Van Kenhove, P. (2008) Consumers’ health-related motive orientations and ready meal consumption behaviour. Appetite 51, 704–712. Griep, L.M.O., Verschuren, W.M.M., Kromhout, D., Ocke, M.C. & Geleijnse, J.M. (2011) Dietary Surveys and Nutritional Epidemiology: colours of fruit and vegetables and 10-year incidence of CHD. Br. J. Nutr. 106, 1562–1569. Grunert, K.G. (1995) Food quality: a means–end perspective. Food Qual. Prefer. 6, 171–176. Grunert, K.G. (2002) Current issues in the understanding of consumer food choice. Trends Food Sci. Technol. 13, 275–285. Kala, A. & Prakash, J. (2006) The comparative evaluation of the nutrient composition and sensory attributes of four vegetables cooked by different methods. Int. J. Food Sci. Technol. 41, 163–171. Moskowitz, H.R. (1995) Food quality: conceptual and sensory aspects. Food Qual. Prefer. 6, 157–162. Poelman, A.A.M. & Delahunty, C.M. (2011) The effect of preparation method and typicality of colour on
ª 2014 The British Dietetic Association Ltd.
R. Bongoni et al.
children’s acceptance for vegetables. Food Qual. Prefer. 22, 355–364. Seljasen, R., Kristensen, H.L., Lauridsen, C., Wyss, G.S., Kretzschmar, U., Birlouez-Aragone, I. & Kahl, J. (2013) Quality of carrots as affected by pre- and postharvest factors and processing. J. Sci. Food Agric. 93, 2611–2626. Simunaniemi, A.M., Nydahl, M. & Andersson, A. (2013) Cluster analysis of fruit and vegetable-related perceptions: an alternative approach of consumer segmentation. J. Hum. Nutr. Diet. 26, 38–47. Verkerk, R., Schreiner, M., Krumbein, A., Ciska, E., Holst, B., Rowland, I., De Schrijver, R., Hansen, M., Gerh€auser, C., Mithen, R. & Dekker, M. (2009) Glucosinolates in Brassica vegetables: the influence of the food supply chain on intake, bioavailability and human health. Mol. Nutr. Food Res. 53, S219. Voorrips, L.E., Goldbohm, R.A., Verhoeven, D.T., van Poppel, G.A., Sturmans, F., Hermus, R.J. & van den Brandt, P.A. (2000) Vegetable and fruit consumption and lung cancer risk in the Netherlands Cohort Study on diet and cancer. Cancer Causes Control 11, 101–115. Wang, W.C., Worsley, A. & Hodgson, V. (2013) Classification of main meal patterns – a latent class approach. Br. J. Nutr. 109, 2285–2296.
ª 2014 The British Dietetic Association Ltd.
Consumer behaviour towards vegetables
Weegels, M.E. & van Veen, M.P. (2001) Variation of consumer contact with household products: a preliminary investigation. Risk Anal. 21, 499–511. Yuan, G.F., Sun, B., Yuan, J. & Wang, Q.M. (2009) Effects of different cooking methods on health-promoting compounds of broccoli. J. Zhejiang Univ. Sci. B 10, 580–588.
Supporting information Additional Supporting Information may be found in the online version of this article: Figure S1. Depicting the water level used to boil the vegetables (a) broccoli and (b) carrots. Low: indicates that a small layer of water was used to boil vegetables; High: indicates that more than a small layer of water was used. Table S1. Chi-squared association between the cluster membership and demographic input variables for broccoli and carrots. Table S2. Chi-squared association between the cluster membership and input variable for broccoli and carrots.
225
PUBLIC HEALTH NUTRITION AND EPIDEMIOLOGY Consumer behaviour towards vegetables: a study on domestic processing of broccoli and carrots by Dutch households R. Bongoni, R. Verkerk, M. Dekker & L. P. A. Steenbekkers Food Quality and Design, Wageningen University, Wageningen, the Netherlands
Keywords cluster analysis, domestic processing, health, selfreporting questionnaire, taste, texture. Correspondence R. Bongoni, Food Quality and Design, Wageningen University, Bornse Weilanden 9, 6708WG, the Netherlands. Tel.: +31 317 48 25 20 E-mail: [email protected] How to cite this article Bongoni R., Verkerk R., Dekker M. & Steenbekkers B. (2014) Consumer behaviour towards vegetables: a study on domestic processing of broccoli and carrots by Dutch households. J Hum Nutr Diet. 28, 219–225 doi: 10.1111/jhn.12245
Abstract Background: Preferences for sensory properties (e.g. taste and texture) are assumed to control cooking behaviour with respect to vegetables. Conditions such as the cooking method, amount of water used and the time-temperature profile determine the nutritional quality (e.g. vitamins and phytochemicals) of cooked vegetables. Information on domestic processing and any underlying motives can be used to inform consumers about cooking vegetables that are equally liked and are nutrient-rich. Methods: Two online self-reporting questionnaires were used to identify domestic processing conditions of broccoli and carrots by Dutch households. Questions on various aspects of domestic processing and consumer motives were included. Descriptive data analysis and hierarchical cluster analysis were performed for both vegetables, separately, to group consumers with similar motives and behaviour towards vegetables. Results: Approximately 70% of consumers boiled vegetables, 8–9% steamed vegetables, 10–15% stir fried raw vegetables and 8–10% stir fried boiled vegetables. Mainly texture was used as a way to decide the ‘doneness’ of the vegetables. For both vegetables, three clusters of consumers were identified: texture-orientated, health-orientated, or taste-orientated. The texture-orientated consumers are identified as the most prevalent (56–59%) group in the present study. Statistically significant associations are found between domestic processing conditions and clusters, whereas no such association are found between demographic details and clusters. Conclusions: A wide variation in domestic processing of broccoli and carrots is found in the present study. Mainly sensory properties (i.e. texture and taste) determined the domestic processing conditions. The findings of the present study can be used to optimise cooking to yield vegetables that meet consumer’s specific sensory preference and are higher in nutrients, and as well as to communicate with target consumer groups.
Introduction Vegetables are processed at the domestic level before consumption. Domestic processing brings about a significant change in the sensorial properties and nutritional quality of vegetables (Kala & Prakash, 2006). Depending on the conditions applied, nutritional quality decreases, ª 2014 The British Dietetic Association Ltd.
e.g. amount of nutrients and phytochemicals, although the bioavailability of certain compounds such as carotenes increases (Dekker et al., 2000). Meanwhile, changes in the sensorial properties such as texture and colour are assumed to be controlled by consumers to meet their (and their family’s) sensory preferences. Therefore, it can be hypothesised that motives such as sensory preferences 219
R. Bongoni et al.
Consumer behaviour towards vegetables
towards vegetables and the habits of consumers determine the cooking conditions (Grunert, 1995; Moskowitz, 1995). When such motives and consumer behaviour towards vegetables are known, food technologists can assess optimised cooking that will result in vegetables (products) that are liked equally but at the same time are higher in nutritional quality (Bongoni et al., 2013). However, for process optimisation detailed information on consumer behaviour towards vegetables and the underlying motives must be known. To our knowledge, there is a lack of such information in the literature. In addition to process optimisation, such information can also be used to cluster consumers into homogenous groups comparable in behaviour and motives (Blanchette & Brug, 2005; Demydas, 2011; Simunaniemi et al., 2013; Wang et al., 2013). Two separate online self-reporting questionnaires are used to identify domestic processing conditions for broccoli and carrots among Dutch households. Broccoli and carrots are chosen for the present study because they are among the commonly consumed vegetables in the Netherlands (Voorrips et al., 2000) and differ morphologically from each other. Both vegetables are rich in a variety of phytochemicals such as polyphenols, glucosinolates and vitamin C in broccoli (Verkerk et al., 2009) as well as phytochemicals such as b-carotene, falcarinol and polyphenols in carrots (Griep et al., 2011; Seljasen et al., 2013). Such compounds are reported to exhibit anti-oxidant and anticarcinogenic activity (Voorrips et al., 2000; Verkerk et al., 2009). In the first part of this paper domestic processing of broccoli and carrots applied by consumers is reported. Next, the results of cluster analysis identifying groups of consumers with comparable domestic processing conditions and motives are shown for both vegetables separately. Materials and methods Online self-reporting questionnaire Two separate online self-reporting questionnaires, which were adapted from the validated questionnaire described by Bongoni et al. (2014), were used to identify domestic processing conditions of broccoli and carrots among Dutch households. Thirteen closed-ended questions were used on the cooking method, part/size of the vegetable piece, water level and type of start used when boiling, cooking time, addition of salt and ways to decide the ‘doneness’ of the vegetable. Water level in the pan for boiling is categorised as: low or high water level (see Supporting Information, Fig. S1). The type of start in the study is categorised as: cold or hot start. In a cold start cooking time commences as the vegetables are heated along with the tap water (approximately 22 °C) in a pan. Cooking time in a hot start commences with the addition of the vegetables into the boiling water. Questions related to the motives for some practices 220
(e.g. water level, type of start, addition of salt) were included. Demographic details (e.g. age, gender, family composition, education) were asked at the end of the questionnaire. The web links to the online surveys were spread among the personal contacts and through the database of Food and Biobased Research (FBR), Wageningen UR, the Netherlands to obtain a large number and maximum variation (in terms of geographical location) of respondents. Participants The target group comprised the main meal preparer of Dutch households. In total, the data from 567 consumers on broccoli and the data from 348 consumers on carrot domestic processing conditions and underlying motives were used for the present study. In the case of broccoli, 12 respondents microwaved and three respondents pressure-cooked the vegetables. In the case of carrots, three respondents microwaved, two pressure-cooked the vegetables and 74 ate carrots raw (and consequently answered all the other questions as ‘none of the above’ or ‘notapplicable’). Given the frequency of these cooking methods and given that the questionnaire asked for domestic processing conditions of vegetables, the data from these participants along with the data from incomplete questionnaires were excluded, resulting in 567 respondents for broccoli and 348 for carrots. The characteristics of the study sample are shown in Table 1. Statistical analysis Statistical analysis was performed using IBM SPSS STATISTICS, version 21(IBM Corp., Armonk, NY, USA). Descriptive statistical analysis was performed to summarise the domestic processing of broccoli and carrots. Cluster analysis was performed to identify consumer groups with similar Table 1 Characteristics of the study sample n (%) Demographic variable Gender Male Female Household type Students (20–25 years) One-person household Couple with children Couple without children Elderly (>65 years) Education level Secondary vocational Higher professional University
Broccoli (n = 567)
Carrots (n = 348)
198 (35) 369 (65)
100 (29) 248 (71)
62 60 208 116 121
38 27 131 96 56
(11) (11) (35) (22) (21)
132 (23) 215 (38) 220 (39)
(11) (8) (37) (28) (16)
96 (28) 132 (38) 120 (34)
ª 2014 The British Dietetic Association Ltd.
R. Bongoni et al.
characteristics (Demydas, 2011) (i.e. domestic processing conditions). Hierarchical cluster analysis was performed with Ward’s method at squared Euclidean distance (Simunaniemi et al., 2013) on variables: part of the vegetable piece, cooking method, water level and, type of start for boiling method and cooking time. The number of clusters was decided based on the dendrogram and logical interpretations of the solutions. There was no substantial correlation (P < 0.05) between the input variables. The clusters were described based on domestic vegetable processing conditions and motives. A chi-squared test was performed to determine statistically significant associations between the clusters and domestic processing variables (P < 0.05).
Consumer behaviour towards vegetables Table 2 Summary of the domestic vegetables applied by Dutch households Aspects on domestic processing of vegetables
Clustering consumers based on domestic processing of broccoli Cluster 1: consumers who are ‘hard texture’-orientated This cluster comprises 39% (n = 221) of the consumers from the study sample who are orientated towards having a ‘hard texture’ of the broccoli. ª 2014 The British Dietetic Association Ltd.
of
Details for carrots (%)
Cooking method
Boiling (71) Steaming (9) Stir frying (10) Boiling followed by stir frying (10)
Boiling (69) Steaming (8) Stir frying (15) Boiling followed by stir frying (8)
Part of the vegetable used
Florets without stalks (10)
Whole carrots (61)
Florets with stalks (72) Florets with stalks and sliced stem (18%)
Cut into pieces (39)
Low water level (18)
Low water level (17)
High water level (82)
High water level (83)
Cold (60)
Cold (73)
Hot (40)
Hot (27)
Cooking time
1–5 min (19) 6–10 min (49) 11–15 min (20) 16–20 min (8) >20 min (4)
1–5 min (7) 6–10 min (40) 11–15 min (27) 16–20 min (21) >20 min (5)
Ways to decide doneness
Texture (64)
Texture (65)
Colour (6) Taste (19) Time (11)
Colour (2) Taste (25) Time (8)
Yes (27)
Yes (29)
Water level when vegetables were boiled
Type of start when boiling
Results Domestic processing of vegetables by consumers Table 2 presents the domestic vegetable processing conditions applied by Dutch households. As expected, a wide variation in the domestic processing of vegetables is observed (Table 2). Approximately 70% of consumers in the present study cooked broccoli and carrots mainly by boiling. Stir frying (10% for broccoli and 15% for carrots), boiling followed by stir frying (10% for broccoli and 8% for carrots) and steaming (9% for broccoli and 8% for carrots) were the other cooking methods used for broccoli and carrots. Most (>80%) consumers reported using a high water level to boil their vegetables. A cold start was applied by 60% and 72% of consumers for boiling broccoli and carrots, respectively. Texture was used by approximately 65% of the consumers to determine the ‘doneness’ of the vegetable during cooking. Except for the type of start, no substantial differences were found in domestic processing between the two vegetables.
conditions
Details for broccoli (%)
Ethical approval The research used online self-reporting questionnaires for information on the domestic processing of vegetables. The respondents were free to participate. Because the research is non-invasive and the details of the participants remain undisclosed, this research does not fall within the remit of the Helsinki Declaration. However, this research received ethical approval from the Social Sciences Ethics Committee, Wageningen University, the Netherlands and the project complies with the Netherlands Code of Conduct for Scientific Practice.
processing
Addition of salt
These consumers boiled the broccoli (100%) using a high water level (100%) and either a cold start (53%) or a hot start (47%). The water level and the type of start used were motivated as a habitual practice. Broccoli is cooked for 6–10 min (77%). Texture change during cooking was used to decide on the ‘doneness’ of broccoli (65%) and 28% of consumers from this group added salt. Cluster 2: consumers who are health-orientated This cluster comprises 33% (n = 188) of consumers from the study sample who are orientated towards having higher health benefits as a result of the consumption of broccoli. This cluster constitutes consumers who use different cooking methods: boiling (40%), steaming (27%), stir 221
Consumer behaviour towards vegetables
frying (29%) and boiling followed by stir frying (4%). When methods such as boiling or boiling followed by stir frying were used, a low water level was used (100%) to prevent nutrient losses into the cooking water. A cold start (100%) was used as a habitual practice. Short to medium cooking times: 1–5 min (26%) and 6–10 min (46%) were used by this group of consumers. Again, the texture of broccoli during cooking was used to decide on ‘doneness’ (56%). Compared to the total study sample (Table 2), a lower number of consumers (16%) added salt to the broccoli. Motives for not adding salt were to restrict the salt intake and to perceive the vegetable taste. Cluster 3: consumers who are ‘medium to soft texture’-orientated This cluster comprises 20% (n = 114) of consumers from the study sample who are orientated towards having a ‘medium to soft’ texture of the broccoli. These consumers mostly boiled the broccoli (97%) using a high water level (100%) and a cold start (70%). Again, habit was the reason for applying this water level and the type of start. Broccoli was cooked for 11–15 min (57%) or for 16–20 min (28%), which can result in a ‘medium to soft’ texture of broccoli. Texture of broccoli during cooking was used to decide on ‘doneness’ (66%). In this cluster, 39% of consumers added salt, which is a higher number than in the total sample (Table 2). Cluster 4: consumers who are taste-orientated This cluster comprises 8% (n = 44) of consumers from the study sample who are orientated towards the taste of the broccoli. These consumers used boiling followed by stir frying (100%) to cook the broccoli. When boiling, a high water level was used (100%) as a habit and mainly a hot start was used (69%) to speed-up the cooking process. Cooking times of 1–5 min (30%) or 6–10 min (52%) were used. Besides texture (55%), taste (40%) was used to assess the ‘doneness’ of the broccoli, which is different from the total sample (Table 2). In this cluster, 39% of the consumers used salt for taste reasons, which is a higher number compared to the total sample (Table 2). Clustering consumers based on domestic processing of carrots Cluster 1: consumers who are texture-orientated This cluster comprises 56% (n = 196) of consumers from the study sample who are orientated towards the texture of the cooked carrots. Carrots are cooked by boiling (100%) using a high water level (100%) and a cold start (100%) as a habitual pr222
R. Bongoni et al.
actice. Carrots are cooked over a range of time that can result in ‘hard to soft’ carrots. In this cluster, 35% of the consumers cooked carrots for 6–10 min, 29% for 11– 15 min and 29% for 16–20 min. Texture of carrots was used to decide on ‘doneness’ by 71% of the consumers. Salt was added habitually by 38% of the consumers, which is higher than in the total sample (Table 2). Cluster 2: consumers who are health-orientated This cluster comprises 18% (n = 63) of consumers from the study sample who are orientated towards the health benefits of the consumption of carrots. Carrots were cooked either by boiling (65%), steaming (16%) or stir frying (19%). To boil, a low water level was used to prevent nutrient losses and a cold start was used as a habitual practice. Carrots were cooked for 6–10 min by 28% of the consumers and for 11–15 min by 38% of the consumers. Texture of the carrots was used (67%) to decide on ‘doneness’. A relatively low number of consumers (13%) added salt in comparison to the total sample (Table 2). Motives for not adding salt were to restrict salt intake and to perceive the carrot taste. Cluster 3: consumers who are taste-orientated and want ‘hard texture’ carrots This cluster comprises 18% (n = 63) of consumers from the study sample who are orientated towards the taste of carrots that are ‘hard’ in texture. Consumers of this cluster used different methods to cook carrots: stir frying (69%), steaming (26%) and boiling followed by stir frying (5%). A cooking time of 6– 10 min was used by most of the consumers (69%). Texture (42%) or taste (40%) was used to determine the ‘doneness’ of carrots. This cluster has a higher percentage of salt-users (37%) compared to the total sample (Table 2). Cluster 4: consumers who are taste-orientated and want ‘medium to soft texture’ carrots This cluster comprises 8% (n = 26) of consumers from the study sample who are orientated towards the taste of carrots that are ‘medium to soft’ in texture. These consumers used boiling followed by stir frying (100%) to cook carrots. For boiling carrots, these consumers used a high water level (100%) and either a cold start (53%) or a hot start (47%) as a habitual practice. The carrots were first boiled for 6–10 min (50%) or for 11–15 min (27%), which was followed by stir frying. Texture (46%) or taste (54%) of the carrots was used to decide on ‘doneness’. This cluster has the highest percentage of salt-users (39%) in comparison to the total sample (Table 2).
ª 2014 The British Dietetic Association Ltd.
R. Bongoni et al.
Association between the input variables of the cluster analysis and the clusters The input variables of the cluster analysis (i.e. the domestic processing conditions) and the clusters from the cluster analysis showed a statistically significant association with each other, indicating that these variables and the clusters are related (see Supporting Information, Table S1). While for both vegetables, the demographic details of the respondents did not show a statistically significant association with the clusters (see Supporting Information, Table S2).
Discussion To our knowledge, the present study is the first to investigate consumer behaviour towards vegetables (i.e. processing of vegetables at the domestic level) and to cluster consumers based on their reported behaviour and motives. Most (70%) consumers cooked vegetables by boiling. Boiling compared to steaming results in a higher loss of (water-soluble) phytochemicals by leaching into the cooking water (Kala & Prakash, 2006). Leaching depends on the vegetable-water ratio used and the cooking timetemperature profile (Verkerk et al., 2009). Although steaming is reported to reduce the loss of phytochemicals (Bernhardt & Schlich, 2006; Yuan et al., 2009), only a few consumers (8–9%) in the present study steamed vegetables. Stir frying, where the temperature of cooking reaches approximately to 140 °C, is detrimental to a great extent for phytochemical stability (Yuan et al., 2009). Approximately 10–15% of the consumers in the present study stir fried vegetables. Texture, which is reported to determine the acceptability of foods (Poelman & Delahunty, 2011), is also used as the main indicator to determine the ‘doneness’ of the vegetables in the present study. Weegels & van Veen (2001) reported a wide variation in behaviour between consumers when performing the same household activity. Such wide variation in behaviour between consumers is also observed when processing vegetables at the domestic level. The quality of cooked vegetables at the time of consumption depends on the way that they were processed (Kala & Prakash, 2006) (e.g. at the domestic level). The information obtained on variation in the domestic processing of vegetables will help to identify aspects of domestic processing that might determine the health and sensory properties of cooked vegetables. In the present study, cluster analysis presented three groups of consumers for both vegetables: those who are texture-orientated, health-orientated or taste-orientated. In the case of broccoli, the texture-orientated consumers and ª 2014 The British Dietetic Association Ltd.
Consumer behaviour towards vegetables
in the case of carrots, the taste-orientated consumers were divided further based on the applied cooking time that results in different textures. The texture-orientated consumers (clusters 1 and 3 in the case of broccoli and cluster 1 in the case of carrots) are the most prevalent group comprising 56–59% of the study sample. These clusters of consumers mostly used a high water level with a cold start and often cooked the vegetables longer than the total sample does (i.e. 6– 10 min) (Table 2), except for the ‘hard texture’-orientated consumers of broccoli. Additionally, compared to the total sample, this cluster of consumers consists of a relatively high percentage of salt-users (28–39%). The behaviour of these consumers comprises mostly habitual practices, which could arise as a result of performing a behaviour repeatedly over time (Simunaniemi et al., 2013). Furst et al. (1996) reported that a past behaviour or experience greatly influences the current behaviour of the consumer. The quality of a product that a consumer experienced or gained (Furst et al., 1996) (in this case, a specific texture of a vegetable that was obtained in previous cooking sessions) might have led to a repeated behaviour in the next cooking sessions. This repeated behaviour probably turned into a habit, for which consumers find it difficult to explain details and specific factors that contribute to their behaviour (Brug et al., 1995; Simunaniemi et al., 2013). Based on the prevalence of this group and the processing conditions that can be detrimental to (nutrients and) phytochemicals, these consumers can be assisted with information on optimised processing conditions, which will result in vegetables within the range of an ‘acceptable texture’ but at the same time are higher in nutritional quality. The health-orientated consumers counted for 20–30% of the study sample for both vegetables. Steaming was used by a higher number of consumers in this group: 27% for broccoli and 16% for carrots, compared to the total sample (Table 2). When applicable, a low water level was used to prevent nutrient losses during cooking and shorter cooking times were applied for broccoli and slightly longer for carrots (with respect to the total sample as shown in Table 2), given the morphological difference between vegetables. Fewer consumers used salt to restrict their salt intake. Grunert (2002) and Brunsø et al. (2002) reported that consumers relate healthiness to eating healthy foods and to avoiding unhealthy foods, aiming to fulfil their basic life values such as good health, a long life and family welfare (Geeroms et al., 2008). The taste-orientated consumers (cluster 4 for broccoli and clusters 3 and 4 for carrots) mainly stir fried the raw or boiled vegetables, cooked with a high water level and mainly applied a hot start, all of which can result in high nutritional quality losses of vegetables. Taste was used to 223
Consumer behaviour towards vegetables
assess the ‘doneness’ of vegetables and was also the reason for the use of salt. The limitation of the present study is that the online self-reporting questionnaires used are not checked for reliability and validity, although they are adapted from the reliable and valid self-reporting questionnaire described by Bongoni et al.(2014). Any adapted questionnaire should be validated prior to administration (Barton et al., 2011). In summary, the findings of the present study are novel and can be used (by food technologists) to optimise cooking, which results in vegetables that meet consumers’ sensory preferences (e.g. texture) and higher in nutrients (Bongoni et al., 2013). Such information on optimised cooking developed for every consumer cluster identified in the present study will assist these consumers to adapt their existing cooking practices. This cluster-specific information can be communicated to consumers in different ways (e.g. through forums, blogs, recipe books or television programmes or cooking classes and by placing pamphlets in the vegetable shops or supermarkets). Cooking programmes at schools could be organised that shall assist in inculcating ‘healthy cooking habits’ in children. To generalise the consumer clusters found in the present study, further research is recommended to cluster and compare consumer behaviour towards vegetables of the same morphology (e.g. potatoes to carrots or cauliflower to broccoli) and of different morphology (e.g. green leafy vegetables). Acknowledgments The present study was supported financially by the VLAG graduate school, Wageningen University, the Netherlands.
Conflict of interests, source of funding and authorship The authors declare that they have no conflict of interests. No funding declared. RB collected the data, analysed, interpreted and wrote the manuscript. All the authors together designed the study, interpreted the data and contributed to writing and critically reviewing the manuscript submitted for publication.
References Barton, K.L., Wrieden, W.L. & Anderson, A.S. (2011) Validity and reliability of a short questionnaire for assessing the
224
R. Bongoni et al.
impact of cooking skills interventions. J. Hum. Nutr. Diet. 24, 588–595. Bernhardt, S. & Schlich, E. (2006) Impact of different cooking methods on food quality: retention of lipophilic vitamins in fresh and frozen vegetables. J. Food Eng. 77, 327–333. Blanchette, L. & Brug, J. (2005) Determinants of fruit and vegetable consumption among 6–12-year-old children and effective interventions to increase consumption. J. Hum. Nutr. Diet. 18, 431–443. Bongoni, R., Steenbekkers, L.P.A., Verkerk, R., van Boekel, M.A.J.S. & Dekker, M. (2013) Studying consumer behaviour related to the quality of food: a case on vegetable preparation affecting sensory and health attributes. Trends Food Sci. Technol. 33, 139–145. Bongoni, R., Verkerk, R., Dekker, M. & Steenbekkers, L.P.A. (2014) Evaluation of research methods to study domestic food preparation. Br. Food J. [Epub ahead of print]. Brug, J., Lechner, L. & De Vries, H. (1995) Psychosocial determinants of fruit and vegetable consumption. Appetite 25, 285–296. Brunsø, K., Fjord, T.A. & Grunert, K.G. (2002) Consumer’s food choice and quality perception, working paper no. 77, The MAPP Centre: The Aarhus School of Business. Dekker, M., Verkerk, R. & Jongen, W.M.F. (2000) Predictive modeling of health aspects in the food production chain: a case study on glucosinolates in cabbage. Trends Food Sci. Technol. 11, 174–181. Demydas, T. (2011) Consumer segmentation based on the level and structure of fruit and vegetable intake: an empirical evidence for US adults from the National Health and Nutrition Examination Survey (NHANES) 2005–2006. Public Health Nutr. 14, 1088–1095. Furst, T., Connors, M., Bisogni, C.A., Sobal, J. & Falk, L.W. (1996) Food choice: a conceptual model of the process. Appetite 26, 247–266. Geeroms, N., Verbeke, W. & Van Kenhove, P. (2008) Consumers’ health-related motive orientations and ready meal consumption behaviour. Appetite 51, 704–712. Griep, L.M.O., Verschuren, W.M.M., Kromhout, D., Ocke, M.C. & Geleijnse, J.M. (2011) Dietary Surveys and Nutritional Epidemiology: colours of fruit and vegetables and 10-year incidence of CHD. Br. J. Nutr. 106, 1562–1569. Grunert, K.G. (1995) Food quality: a means–end perspective. Food Qual. Prefer. 6, 171–176. Grunert, K.G. (2002) Current issues in the understanding of consumer food choice. Trends Food Sci. Technol. 13, 275–285. Kala, A. & Prakash, J. (2006) The comparative evaluation of the nutrient composition and sensory attributes of four vegetables cooked by different methods. Int. J. Food Sci. Technol. 41, 163–171. Moskowitz, H.R. (1995) Food quality: conceptual and sensory aspects. Food Qual. Prefer. 6, 157–162. Poelman, A.A.M. & Delahunty, C.M. (2011) The effect of preparation method and typicality of colour on
ª 2014 The British Dietetic Association Ltd.
R. Bongoni et al.
children’s acceptance for vegetables. Food Qual. Prefer. 22, 355–364. Seljasen, R., Kristensen, H.L., Lauridsen, C., Wyss, G.S., Kretzschmar, U., Birlouez-Aragone, I. & Kahl, J. (2013) Quality of carrots as affected by pre- and postharvest factors and processing. J. Sci. Food Agric. 93, 2611–2626. Simunaniemi, A.M., Nydahl, M. & Andersson, A. (2013) Cluster analysis of fruit and vegetable-related perceptions: an alternative approach of consumer segmentation. J. Hum. Nutr. Diet. 26, 38–47. Verkerk, R., Schreiner, M., Krumbein, A., Ciska, E., Holst, B., Rowland, I., De Schrijver, R., Hansen, M., Gerh€auser, C., Mithen, R. & Dekker, M. (2009) Glucosinolates in Brassica vegetables: the influence of the food supply chain on intake, bioavailability and human health. Mol. Nutr. Food Res. 53, S219. Voorrips, L.E., Goldbohm, R.A., Verhoeven, D.T., van Poppel, G.A., Sturmans, F., Hermus, R.J. & van den Brandt, P.A. (2000) Vegetable and fruit consumption and lung cancer risk in the Netherlands Cohort Study on diet and cancer. Cancer Causes Control 11, 101–115. Wang, W.C., Worsley, A. & Hodgson, V. (2013) Classification of main meal patterns – a latent class approach. Br. J. Nutr. 109, 2285–2296.
ª 2014 The British Dietetic Association Ltd.
Consumer behaviour towards vegetables
Weegels, M.E. & van Veen, M.P. (2001) Variation of consumer contact with household products: a preliminary investigation. Risk Anal. 21, 499–511. Yuan, G.F., Sun, B., Yuan, J. & Wang, Q.M. (2009) Effects of different cooking methods on health-promoting compounds of broccoli. J. Zhejiang Univ. Sci. B 10, 580–588.
Supporting information Additional Supporting Information may be found in the online version of this article: Figure S1. Depicting the water level used to boil the vegetables (a) broccoli and (b) carrots. Low: indicates that a small layer of water was used to boil vegetables; High: indicates that more than a small layer of water was used. Table S1. Chi-squared association between the cluster membership and demographic input variables for broccoli and carrots. Table S2. Chi-squared association between the cluster membership and input variable for broccoli and carrots.
225
