Sage Journals: Discover world-class research

Abstract

This study aims to address gaps in consumer-centred research on traditional small grain foods in Africa by developing an integrated model for explaining consumer purchase intentions towards traditional small grain foods in Zimbabwe. A survey of 386 Zimbabwean adults between the ages of 18 and 65 years was administered in Zimbabwe’s 10 provinces. Sequential quantitative triangulation using Covariance-Based SEM (CB-SEM) and Partial Least Squares SEM (PLS-SEM) was used for data analysis. The study empirically validated an eleven-factor traditional small grain food preference model, which exhibited acceptable fit, moderate explanatory power and significant predictive power. Implications for academia, policy and practice are discussed.

Keywords

purchase intentions traditional foods model Zimbabwe consumer behaviour sorghum millet rapoko small grains

Introduction

Understanding consumers and their choices, and placing them at the forefront of academic research is crucial for addressing human health and other challenges (Chen & Antonelli, 2020). Consistent with this view, Moyo et al. (2021) advocated for more consumer-centred research on traditional small grain foods to stimulate demand for, and consumption of the same. The uptake of traditional small grain foods is known to have several beneficial outcomes, including the improvement of food and nutritional security for consumers (Banerjee & Maitra, 2020; McCartan et al., 2020), sustenance of livelihoods and incomes for farmers (Phiri et al., 2019), environmentally sustainable agriculture (Akinola et al., 2020), and promotion of sustainable food systems (Akinola et al., 2020). Traditional small grain foods are of particular importance for regions prone to drought, such as Africa, which faces high levels of food and nutritional insecurity and experiences the highest rates of undernourishment globally (Rampa et al., 2020). Through consumer-centred research, there is potential to obtain insights that could inform product marketing strategies and policy interventions aimed at attaining the benefits of traditional small grain food consumption.

While the foregoing discussion accentuates the importance of consumer-centred research in driving consumer demand for traditional food products, literature also identifies significant deficiencies in existing research on consumer-centred research on traditional foods, particularly in Africa, despite the abundance of diverse traditional foods. Firstly, while food-related consumer-centred research is well-established in many regions of the world, there is a notable scarcity of such studies focusing on traditional food products in the African context. For example, a systematic literature review by García-Barrón et al. (2021) found that consumer-centred research on traditional foods was scarce in Africa. Most of these studies were conducted in European countries like Italy, Spain, and France. This scarcity of African studies is well illustrated by Moyo et al. (2021) who established that from the year 2000–2021, only 33 empirical studies were published on consumer behaviour relating to African traditional foods. In addition, these 33 empirical studies were conducted in only 14 of the 54 African countries, which demonstrates the absence of such research in most African countries. The limited consumer behaviour research on traditional foods in Africa hinders the development of effective strategies to promote the consumption of traditional foods. As a result, the potential benefits associated with traditional food consumption become more difficult to realise.

Secondly, existing consumer behaviour research on traditional foods in Africa has tended to explore a limited range of factors. For instance, Moyo et al. (2023) found that 59% of the research topics on traditional foods in Africa focused on sensory food attributes, much to the neglect of the wide spectrum of possible factors affecting consumer behaviour. The multiple conscious and unconscious factors influencing consumer behaviour are complex (Shen & Chen, 2020) and necessitate a more comprehensive examination of factors.

Finally, existing consumer-centred studies on traditional food consumption in Africa have often lacked a strong theoretical foundation. Moyo et al. (2023) found that 85% of consumer-centred African studies did not refer to any theory. The limited use of consumer behaviour theories to guide research in this field may be explained by the facts that several of those studies were conducted in areas outside of consumer behaviour or business, such as biological sciences (Afolabi et al., 2020), crop science (Bechoff et al., 2018), nutrition and paediatrics (Coley et al., 2020), and agriculture and food science (de Beer et al., 2016). For the few studies that did reference consumer behaviour theories, none comprehensively applied a selected theory or model. This means that most consumer-centred research on African traditional foods has not fully benefitted from established consumer behaviour theory. But more importantly, consumer behaviour theory has in turn not benefitted from existing African research as there has been no significant efforts to apply complete theories, integrate multiple theories and consolidate existing fragmented knowledge into a holistic model for explaining and possibly predicting the purchase and consumption of traditional foods by in an African context (Moyo et al., 2023).

The identified deficiencies constrain the development of marketing theory in Africa. Consequently, marketing practitioners are deprived of insights and tools essential for the achievement of a demand-led growth and sustainability of the traditional food market in Africa. The point of departure for resolving this constraint is the development of marketing theory. To this end, this research aims to enhance marketing theory by creating a comprehensive model that considers various factors and how they interact to predict consumer intent to purchase traditional small grain foods in Zimbabwe. Accordingly, the objective of the study is to propose an integrated model to guide marketing practitioners and retailers on how to design marketing interventions that stimulate consumer purchase intentions towards traditional small grain foods.

Literature review

Traditional small grain foods in Zimbabwe

Traditional foods include a variety of fruits, grains, pulses and tubers (Akinola et al., 2020). Among these, ‘small grains’, which are part of the grass family, feature prominently, with sorghum, millet and rapoko being the most widely cultivated in Zimbabwe (Phiri et al., 2019). The most cultivated type of millet in Zimbabwe is pearl millet, hence the term ‘millet’ is generally understood to mean pearl millet. Rapoko is a type of ‘minor millets’ and is sometimes called ‘finger millet’ (Hassan et al., 2021). However, the term ‘Rapoko’ is the more commonly used name in Zimbabwe. Sorghum and millet are rich in compounds that provide a variety of health benefits, including antimicrobial, antilipidemic, anticancer, anti-inflammatory, antiproliferative and certain antidiabetic effects (Senevirathne et al., 2021).

The cultivation of sorghum, millet and rapoko serves multiple purposes, including human consumption, animal feed (Gupta et al., 2023) and recreational uses, such as brewing traditional opaque beers at both household and industrial levels (Jane et al., 2015). A significant portion of these grains is cultivated by smallholder farmers primarily for subsistence (Kudita et al., 2023). The grains are typically harvested and stored for family use, with households incorporating them into various traditional dishes, including porridge and fermented drinks like “Mahewu” (Sheethal et al., 2022). In addition to home production, traditional small grains can also be found in local shops and food markets, providing consumers with options to purchase them in different forms, such as flour or whole grains (Saleh et al., 2013).

For centuries, traditional crops have served as the primary staples in Africa, prior to the rise of maize as a cash crop (Muthamilarasan & Prasad, 2021). In the early to mid-20th century, traditional grains were staple foods for many Zimbabwean households, especially smallholder farmers who depended on these drought-resistant crops for food security (Ndlovu et al., 2020; Nyasha et al., 2023). However, following Zimbabwe’s independence in 1980, the agricultural focus shifted towards commercial maize production, which was seen as a more dependable and financially rewarding staple (Chopera et al., 2022; Tanyanyiwa, 2021).

By the late 20th century, there was a significant decline in the consumption of traditional grains, which coincided with a nutritional transition that favoured processed foods like wheat and rice products (Chopera et al., 2022). This change was intensified by socio-economic developments, urbanisation and the promotion of maize as the main staple, which led to reduced cultivation and consumption of sorghum, millet and rapoko (Chopera et al., 2022; Kauma & Swart, 2022). Government policies further reinforced maize production, often to the neglect of traditional grains (Kauma & Swart, 2022; Tanyanyiwa, 2021).

Recently, there has been a renewed interest in traditional grains, driven by a growing awareness of their nutritional advantages and resilience to climate change. This renewed interest has been bolstered by initiatives aimed at promoting food diversity and improving food security amid climate variability (Mushore et al., 2021). There is rising interest in Sub-Saharan Africa for substituting maize with climate-smart crops like sorghum and millets in local food processing. This study, through its comprehensive consumer-focused approach, contributes to the aforementioned endeavour by developing an understanding of the factors affecting consumer purchase intentions toward traditional small grain foods.

Theoretical background

In exploring factors influencing consumer purchase intentions towards traditional small grain foods, this study draws upon three seminal models - the Food Preference Model (Randall & Sanjur, 1981), the Consumer Behaviour Model with respect to food (Steenkamp, 1997), and the Theory of Planned Behaviour (Ajzen, 1991). Each model offers valuable perspectives for a comprehensive understanding of factors influencing consumer purchase intentions toward traditional small grain foods.

The Food Preference Model introduced by Randall and Sanjur (1981) is one of the earliest frameworks that considers various attributes affecting food choice. It proposes that food consumption is influenced by food preference, which in turn is shaped by characteristics of the individual (e.g. demographics, attitudes), the food itself (e.g. sensory appeal, cost) and the environment (e.g. household size, culture). While comprehensive, this model is primarily descriptive and lacks the mechanisms to quantify and predict the relative impact of different factors (Gorton & Barjolle, 2013).

Building on Randall and Sanjur (1981) model, Steenkamp’s (1997) developed the Consumer Behaviour Model with respect to food, which introduces a four-step decision-making process to explain food preferences and choices. This model retains the individual, food and environmental factor categories from Randall and Sanjur (1981), but emphasises the underlying decision process. However, like its predecessor, the Consumer Behaviour Model with respect to food stops short of identifying causal relationships, thereby limiting its explanatory and predictive power (Gorton & Barjolle, 2013).

The Theory of Planned Behaviour (Ajzen, 1991) addresses this lack of explanatory and predictive power by asserting that behaviour is directly influenced by intentions and perceived behavioural control. The theory further asserts that intentions are shaped by attitudes, subjective norms and perceived control. While the TPB enables the quantitative assessments of relationships, it does not encompass as many factors as the food preference models and tends to focus on rational decision-making, often overlooking product-specific variables (Armitage et al., 1999; Chen, 2012). Ajzen (2020) acknowledges the benefit of extending his model by incorporating additional predictors.

The reviewed models confirm that no single model fully captures the range of food-specific factors and their interactions in a measurable way, which would enable the prediction of consumer purchase intentions towards traditional small grain foods in Zimbabwe. This study aims to fill that gap. While the Food Preference Model (Randall & Sanjur, 1981) and the Consumer Behaviour Model with respect to food (Steenkamp, 1997) identify various factors that influence consumer food choices, they lack predictive and explanatory power as they cannot measure those identified factors (operability). Similarly, although the TPB offers strong explanatory and predictive power, it does not address as many factors as the Food Preference Model with respect to food and the Consumer Behaviour Model, particularly regarding product attributes.

By integrating these models, we seek to develop a comprehensive model that encompasses a wider range of factors while remaining measurable and operable. By integrating core tenets from the Food Preference Model, Consumer Behaviour Model and Theory of Planned Behaviour, a more holistic framework emerges.

Conceptual framework

In developing a comprehensive model to identify factors influencing traditional small grain food purchase intentions, this study integrated the Theory of Planned Behaviour (Ajzen, 1991), Consumer Behaviour Model with respect to food (Steenkamp, 1997), and Food Preference Model (Randall & Sanjur, 1981). Following guidelines by Ajzen (2010), only directly causal factors from the above reference models were included in the conceptual framework. Descriptive processes like the decision-making steps and background factors like demographics were excluded from the conceptual framework.

The resulting conceptual framework (Figure 1) identifies ten independent factors categorised into three categories: external factors, food properties and personal factors. This integrative conceptual framework combines the TPB’s robust predictive mechanisms with the expansive pool of potential influencing factors proposed in the Consumer Behaviour Model with respect to food (Steenkamp, 1997) and Food Preference Model (Randall & Sanjur, 1981).

Figure 1.

Conceptual framework. Source: Adapted from Steenkamp (1997), Ajzen (1991) and Randall and Sanjur (1981).

Factors influencing food purchase intentions

Numerous factors influence consumer purchase intentions towards traditional small grain foods as shown in Figure 1 above. Physiological needs based on the body’s metabolism and nutritional requirements play a fundamental role (Steenkamp, 1997; Williams, 2021). The satiating nature of whole grains makes them desirable from a physiological perspective (Penhill, 2020). Similarly, research indicates that sensory food properties like appearance, flavour, aroma and texture significantly impact willingness to purchase foods like nuts (Hong et al., 2020).

Price, as an indication of affordability, is another driver of purchase decisions for organic foods (Chauke, 2018). High prices have been found to deter consumption if perceived benefits do not outweigh costs (Doan, 2021; Hong et al., 2020). Cultural identity and traditional foods’ role in establishing local heritage are influential factors, with foods bearing symbolic cultural meanings (Frez-Muñoz et al., 2021; Hansena et al., 2018). Prior research suggests identity, in the broadest sense, can impact behaviour intentions and actual behaviour (Conner & Armitage, 2002).

Promotion through advertising, personal selling, and other marketing communications, for instance, is influential in raising awareness, altering attitudes and promoting purchases (Chauke, 2018; Kung et al., 2021). Inadequate promotion of foods products by manufacturers or retailers has previously been identified as a barrier for local food purchases (Garbacz, 2017). Availability, or perceived convenience of accessing preferred foods near home impacts choices. Empirical evidence suggests that low availability diminishes purchase intentions even for desired products (Chauke, 2018; Doan, 2021; Loh & Hassan, 2022).

Subjective norms, reflecting perceived social pressure from important referent groups like family, friends and society, also influences purchase intentions (Ajzen, 1991; Shen et al., 2022). The family is considered the most vital reference group influencing consumer values, attitudes and decisions (Blythe, 2013; Hawkins & Mothersbaugh, 2013).

Perceived behavioural control, based on perceptions of resources, opportunities and obstacles to performing a behaviour, impacts intentions - greater perceived control enhances intentions (Ajzen, 1991; Albarracín et al., 2016; Joshi et al., 2021). Prior research confirms perceived control positively influences consumer beliefs about ability to undertake behaviours.

Attitude towards a behaviour stemming from beliefs about associated outcomes is predictive of intentions (Ajzen, 1991; Chen & Deng, 2016; Hsu, 2014). Empirical studies confirm that attitudes explained substantial variance in organic food purchase intentions (Ercis et al., 2020; Zayed et al., 2022).

Another factor is the risk of non-communicable diseases, which is an aspect of health consciousness motivating healthier food choices like organic purchases (Choi et al., 2021; Jeong & Kim, 2020; Verbeke, 2005). On a broader level, previous research identified health consciousness as an important psychological factor influencing food choice.

Purchase intention, the sole dependent factor of the conceptual framework, is the willingness of consumers to buy products or services (Ahmad et al., 2019), reflecting motivation to shop and the likelihood of making a purchase (Fang, 2022). It serves as a precursor to actual purchase behaviour (Ajzen, 2020), providing insights into consumer preferences (Thilina & Gunawardane, 2019). Empirical research shows that purchase intentions strongly predict purchasing decisions across different contexts, such as environmental purchasing behaviour (Sinha & Annamdevula, 2022), halal food products (Novitasari et al., 2021) and organic food (Ferreira & Pereira, 2023). Understanding purchase intentions is therefore crucial for informing marketing strategies, including product development, pricing, promotion, and distribution decisions, to align with consumer expectations and drive desired behaviours.

Methodology

Research strategy and sampling

This study adopted a survey research strategy to empirically investigate factors influencing purchase intentions for traditional small grain foods among Zimbabwean consumers. Data were collected using a structured face-to-face researcher-administered questionnaire. The target population comprised 7,814,779 (Review World Population, 2022) adult consumers and non-consumers of traditional small grain foods, who are Zimbabwean residents between the ages of 18 and 65 years. Individuals under the age of 18 are minors and were therefore excluded from the study, as parental consent would be required for their participation. Additionally, individuals over the age of 65 were considered vulnerable and excluded from the study, in compliance with guidelines from the board responsible for granting ethical clearance. The designation of research subjects older than 65 years as a vulnerable group is consistent with earlier conclusions by López-Parra et al. (2022).

While the study used a non-probability sample, a probability sample size calculator was used to approximate the minimum sample size of 385 (N = 7 814 779, confidence level = 95% and margin of error = 5%). Not only was this sample size approximation consistent with Saunders et al. (2019)’s assertion that quota sampling requires similar sample sizes to probability samples; but it produced a higher minimum sample size compared to that recommended by alternative nonprobability sample size calculators such as G-Power^® software and the a-priori sample size (calculator) for structural equation models (Soper, 2020). This likely reduced the magnitude of sampling error.

While the minimum approximated sample size was 385, a total of 386 usable questionnaires were obtained and included in the data analysis. With no comprehensive sampling frame, non-probability quota sampling was utilised. Quotas were set for each of Zimbabwe’s 10 provinces based on percentage contribution to the national population to enhance representativeness. Within each province, shoppers exiting three leading supermarkets and a traditional food market were conveniently selected. These supermarkets have not been names for ethical reasons.

The supermarket chains were chosen due to their status as the leading formal distribution entities within Zimbabwe’s concentrated oligopolistic food retail industry (Maumbe & Chikoko, 2020). The decision to include food markets was based on the recognition that they play a vital role in the food supply chain and particularly cater to the under-resourced segments of the community (Muvhuringi et al., 2021).

Data collection instrument

The questionnaire began with a cover letter explaining that the purpose of the research was to identify and assess context-specific factors affecting consumer purchase intentions towards traditional small grain foods (e.g. Sorghum, millet and rapoko) in Zimbabwe. It also explained that participation in the study was voluntary; responses would be kept confidential; and participants could withdraw from the study at any time without being penalised.

The second section gathered the respondents’ demographic information such as province or residence, gender, age range, education, whether they eat traditional small grain foods or not, and monthly income. The last section elicited responses to measurement items for factors on the conceptual framework. A 5-point Likert scales ranging from “strongly disagree” to “strongly agree” was used. Questionnaire items were either adopted or adapted from previously validated scales. A list of assigned codes for the measurement items, the full measurement statement for each item, the source from which the item was either adopted or adapted, and the context in which the measurement items were first utilised in the identified source, are provided in the supplementary material.

Data analysis

Ethical considerations in this study included privacy of participants, voluntary participation and the right to withdraw from the study at any stage, informed consent, confidentiality of data and anonymity of participants (Saunders et al., 2019). The study was granted ethical approval by the ethics board of the associated institution.

Data analysis

This study utilised both covariance-based structural equation modelling (CB-SEM) and partial least squares structural equation modelling (PLS-SEM) as complementary methods, an approach that provides quantitative complementary triangulation. SPSS v26 software was used for CB-SEM. Once the structural model was validated using SPSS, the explanatory and predictive quality of the model was assessed using SmartPLS4, which is a PLS-SEM based analysis software.

Results

Table 1 compares the distribution of 386 respondents across provinces with the corresponding population figures. The sample proportions closely mirror the population distribution, indicating a high representativeness of the provincial spread. For instance, Harare’s sample proportion (16.1%) nearly matches its population share (16.0%), while Mashonaland Central is represented at 9.1%, exactly reflecting its population percentage. Minor discrepancies appear in provinces such as Matabeleland North and South where the sample percentages (5.4% and 4.9%, respectively) slightly underrepresent the population shares (5.5% and 5.0%). Overall, the sample’s provincial composition aligns well with the population structure, suggesting that the findings could be generalised with reasonable confidence across provinces.

Table 1.

Province of respondents.

Province	Sample frequency	Sample percent (%)	Population frequency	Population percent (%)
Manicaland	52	13.5	2,037,703	13.4
Mashonaland central	35	9.1	1,384,891	9.1
Mashonaland East	44	11.4	1,731,173	11.4
Mashonaland west	48	12.4	1,893,584	12.5
Matabeleland North	21	5.4	827,645	5.5
Matabeleland south	19	4.9	760,345	5.0
Midlands	46	11.9	1,811,905	12.0
Masvingo	42	10.9	1,638,528	10.8
Harare	62	16.1	2,427,231	16.0
Bulawayo	17	4.4	665,952	4.4
Total	386	100.0	15,178,957	100.00

Source: Compiled from statistical results and ZIMSTATS 2023.

Validity assessment

The convergent validity of the measurement instrument was assessed through confirmatory factor analysis (CFA). This involved evaluating the standardised factor loadings and average variance extracted (AVE) statistics (Hair et al., 2014). The criteria for validity is that the standardised factor loading of each measurement item must be at least 0.5, and the AVE scores for each factor must be at least 0.5 (Hair et al., 2014).

In the CFA, six out of the 60 measurement items (F2, F3, F7, K1, I3, and K2) exhibited standardised regression weights below the 0.5 threshold. Consequently, these items were excluded from the final factor structure, in line with the decision rule. The remaining 54 items all demonstrated standardised factor loadings exceeding 0.5, thus satisfying the criteria for establishing convergent validity (Hair et al., 2014).

In addition, all constructs demonstrated acceptable validity, as indicated by AVE scores that were all above the minimum required score of 0.5 (Hair et al., 2014), as shown in Table 2. The final list of measurement items, after item deletion, is provided in the supplementary material. These results confirm the validity of the refined measurement instruments.

Table 2.

Composite reliability (CR), Cronbach alpha and average variance extracted (AVE) scores for model constructs.

Factor	Scale item	Standardized regression weights	p-values	CR	Cronbach alpha	AVE	Final number of items
Physiological food properties	B1	0.764	***	0.86	0.86	0.51	6
	B2	0.772	***
	B3	0.755	***
	B4	0.501	***
	B5	0.776	***
	B6	0.663	***
Sensory food properties	C1	0.88	***	0.85	0.87	0.53	5
	C2	0.818	***
	C3	0.555	***
	C4	0.669	***
	C5	0.679	***
Price	D1	0.832	***	0.90	0.90	0.62	6
	D2	0.58	***
	D3	0.548	***
	D4	0.893	***
	D5	0.863	***
	D6	0.918	***
Cultural identity	E1	0.877	***	0.88	0.89	0.55	6
	E2	0.842	***
	E3	0.707	***
	E4	0.82	***
	E5	0.505	***
	E6	0.63	***
Promotion	F1	0.554	***	0.83	0.82	0.56	4
	F4	0.918	***
	F5	0.882	***
	F6	0.556	***
Availability	G1	0.84	***	0.94	0.94	0.75	5
	G2	0.883	***
	G3	0.903	***
	G4	0.868	***
	G5	0.834	***
Subjective norms	H1	0.661	***	0.89	0.89	0.63	5
	H2	0.885	***
	H3	0.893	***
	H4	0.771	***
	H5	0.723	***
Perceived behavioral control	I1	0.728	***	0.90	0.90	0.60	6
	I2	0.788	***
	I4	0.774	***
	I5	0.675	***
	I6	0.841	***
	I7	0.844	***

Factor	Scale item	Standardized regression weights	p-values	CR	Cronbach alpha	AVE	Final number of items
Attitude	J1	0.846	***	0.92	0.91	0.69	5
	J2	0.67	***
	J3	0.92	***
	J4	0.893	***
	J5	0.794	***
Threat of Noncommunicable Diseases	K3	0.551	***	0.80	0.79	0.59	3
	K4	0.978	***
	K5	0.717	***
Purchase intention	L1	0.779	***	0.83	0.79	0.55	4
	L2	0.505	***
	L3	0.818	***
	L4	0.818	***

Source: Compiled from statistical results.

Reliability assessment

Cronbach’s alpha and composite reliability (CR) were used for assessing the reliability of the measurement instrument. The decision rule is that Cronbach’s alpha and CR statistics must be at least 0.7 (Hair et al., 2021) for reliability to be established. Cronbach’s alpha and CR for all constructs met or exceeded the recommended minimum criterion of 0.7 (Hair et al., 2021), thereby indicating that the measurement items used to represent the constructs had satisfactory reliability. Table 2 summarises results obtained for the measures of reliability and validity.

Assessment of measurement model fit

The fit of the measurement model to the survey data was evaluated using various model fit indices to determine how well the model, derived from literature, explained the observed data. Table 3 below presents results obtained for each of the assessed model fit indices and the corresponding minimum values for establishing adequacy of fit. While chi-square, Comparative Fit Index (CFI) and Goodness-of-Fit Index (GFI) did not meet the thresholds for adequate fit, a marginally greater number of indices reached their corresponding thresholds of adequate model fit, namely normed chi-square, Root Mean Square Error of Approximation (RMSEA), Standardized Root Mean Square Residual (SRMR) and Parsimonious Normed Fit Index (PNFI). Based on the predominant results obtained, the hypothesised measurement model demonstrated adequate fit to the collected data.

Table 3.

Measurement model fit indices.

Fit index	Threshold	Source	Obtained value
χ² and parameters significant	Non-significant chi-square	(Hair et al., 2014, pp. 579, 584)	χ² of 4277.860 was significant
χ²/ d.f. Ratio	<3 good fit <5 acceptable fit	(Hair et al., 2014, p. 579)	3.152
RMSEA	Between 0.05 – 0.08	(Hair et al., 2014, p. 577)	0.075
CFI	>0.9	(Bentler, 1990)	0.827
SRMR	0.08≦	(Hu & bentler, 1995)	0.0839
GFI	>0.9	(Hair et al., 2014, p. 579)	0.648
PNFI	>>0.5 or 0.60–0.90	(Dash & Paul, 2021, p. 3; Putra et al., 2021:1479)	0.701

Source: Compiled from statistical results.

Model fit may also be calculated using the formula $G o F = \sqrt{A v e r a g e A V E \times A v e r a g e R^{2}}$ (Pramana & Fachrunnisa 2024). The GoF calculation yielded a value of 0.598, which demonstrates good measurement model fit as it is above the threshold of 0.3 (Chauke & Duh, 2019), thereby confirming conclusions of acceptable measurement model fit based on fit indices assessed above.

Assessment of structural model

Following the validation of the measurement model above, this section examines the quality of the structural model by assessing the model’s quality using covariance-based model fit indices. Additionally, the model’s explanatory and predictive properties, which are additional and complementary quality parameters, are assessed based on the PLS-SEM approach.

Assessment of structural model fit using covariance-based indices

To evaluate model fit, Hair et al. (2014:651) suggest that researchers should at least assess and report one absolute index, one incremental index and the model’s χ2. Furthermore, Ockey and Choi (2015) advise that it is essential to report one parsimonious index as well. Rather than evaluating just one of each type of index mentioned above, this study examines two for each category.

Specifically, it assesses the Root Mean Square Error of Approximation (RMSEA) and the Goodness-of-Fit Index (GFI) as absolute fit indices; the Incremental Fit Index (IFI) and the Comparative Fit Index (CFI) as incremental fit indices; the Parsimonious Normed Fit Index (PNFI) and the Parsimony Comparative Fit Index (PCFI) for parsimonious fit indices; and the Normed Chi-square (χ2/df) along with the associated p-value for the chi-square category.

Table 4 below shows the assessed structural model fit indices and how they compare with generally accepted cut-off values. As shown in Table 4, the normed chi-square value (chi-square/df) was 4.076, which fell within the acceptable upper limit of <5. However, the p-value was significant, when it ought to have been non-significant to achieve adequate model fit (Dash & Paul, 2021; Hair et al., 2014). The RMSEA value of 0.089 indicated adequate structural fit as it was less than the recommended upper limit of 0.1 (Kim et al., 2016). Both the lower bound and upper bound RMSEA values of 0.087 and 0.092, respectively, were within the upper limit of 0.1, confirming the acceptable fit of the model.

Table 4.

Model fit estimates of structural model.

1	Chi-square (χ²)	Cutt-off	Observed value
	Chi-square	-	5715.12
	Degrees of freedom	-	1402
	Chi-square/df	<3 or <5	4.076
	p-value	p > .05	0.000
2	Absolute fit of measure	Cutt-off	Observed value
	Root mean square error of approximation (RMSEA)	<0.8 or <0.1	0.089 (LO CI = 0.087;HI CI = 0.092)
	GFI	>0.95 or >0.90	0.598
3	Incremental fir indices	Cutt-off	Observed value
	Incremental fit index (IFI)	>0.95	0.745
	Comparative fit index (CFI)	>0.95	0.744
4	Parsimonious fit index	Cutt-off	Observed value
	PNFI	>0.5	0.65
	PCFI		0.703

Source: Compiled from statistical results.

The GFI, IFI, and CFI values were well below the recommended threshold. However, both parsimonious model fit indices, PNFI and PCFI demonstrated good model fit as both values were above the cut-off of 0.5. Overall, the hypothesised structural model demonstrated adequate model fit for half of the assessed indices (four out of eight), suggesting a fair level model fit. Figure 2 below presents a visual depiction of the structural model assessed by means of the fit indices discussed above.

Figure 2.

Construct Structural model. Source: Compiled based on statistical data from the study.

In addition to the use of CB-SEM based structural model fit indices, the quality of the model was also analysed by assessing the model’s explanatory and predictive power, as recommended by Hair et al. (2021). This approach is consistent with earlier guidance by Anderson and Gerbing (1988), who asserted that fit may be evaluated based on the percentage of variance explained in the specified regressions, which is basically a model’s explanatory power. Assessment of these additional aspects of model quality were performed using Smart-PLS 4 and the results are presented below. Having evaluated the quality of the structural model using covariance-based model fit indices, the following sections assess the explanatory and predictive capabilities of the structural model through the PLS-SEM approach.

Assessment of model explanatory power

The explanatory power of the structural model was assessed by examining the coefficient of determination (R²), which represents the proportion of endogenous construct variance explained by all the exogenous constructs linked to it (Ismael & Duleba, 2021). According to Hair et al. (2021), R² values of 0.75, 0.50, and 0.25 indicate considerable, moderate, and low levels of explanatory power, respectively. The R² for purchase intentions was 0.595 as obtained from SmartPLS4 results, showing a moderate predictive power. This implies that 60% of the variance observed in purchase intentions could be explained by the ten independent variables of the conceptual framework.

The effect size (f²), which illustrates the magnitude by which one exogenous factor contributes to explaining an endogenous factor in terms of R², was also assessed for each independent factor. An f² value below 0.02 denotes a small effect, a value between 0.02 and 0.15 signifies a moderate effect and anything above 0.15 indicates a strong effect (Ismael & Duleba, 2021). Attitude had the greatest effect size (f² = 0.103) on R², followed by price (f² = 0.035) and cultural identity (f² = 0.033), which may be classified as a moderate effect. The path with the least impact was threat of non-communicable diseases (f2 = 0.001), promotion (f² = 0.003), physiological food properties (f² = 0.008) and perceived behavioural control (f² = 0.011), which may be classified as a small effect size. In summary, six out of the ten paths demonstrated moderate effect sizes (f² > 0.002) while the remaining four displayed a small effect.

Assessment of model predictive power

The predictive performance of the structural model was assessed using predictive relevance (Q²). The predictive relevance (Q²) values were all greater than 0, indicating acceptable predictive power of the model (Fernández et al., 2018). The PLS_predict algorithm’s output was interpreted following guidelines by Shmueli et al. (2019), which suggested that the model exhibited a low out-of-sample predictive power.

Discussion

Attainment of research objectives

Psychometric analysis of the measurement instrument demonstrated positive results as illustrated by a Cronbach’s alpha and composite reliability (CR) that were greater than 0.70. Additionally, all factors achieved acceptable average variance extracted (AVE) levels by surpassing the minimum threshold score of 0.5 (Hair et al., 2014). Convergent validity criterion was met as all retained measurement items had a minimum standardised factor loading of 0.5.

The main objective was to propose an integrated model to assist marketing professionals and retailers in crafting strategies that stimulate consumer purchase intentions for traditional small grain foods. This objective was attained by empirically testing a conceptual framework, resulting in a validated Traditional Small Grain Food Preference Model. Empirical evidence demonstrated that the model is acceptable based on the structural model fit indices. Specifically, the model met 4 out of the 8 overall fit indices that were evaluated. In terms of index classification, the model satisfied at least 50% of the Chi-Square (X²), absolute fit and parsimonious fit indices. Only the incremental category of fit indices had less than half of the examined indices meeting the minimum acceptable thresholds.

Using the coefficient of determination (R²) as an indicator of explanatory power, the model exhibited moderate explanatory power as the 10 independent factors could account for about 60% of the observed variation in purchase intentions (R² = 0.595). The model also demonstrated significant predictive power, though it was relatively low.

Contribution of the study

This study makes a significant theoretical contribution to the existing body of knowledge on consumer behaviour concerning traditional small grain foods in Zimbabwe. The study’s empirical results offer new insights that extend the boundaries of current knowledge for consumer behaviour scholars.

This study produced an empirically validated model that explains and predicts consumer purchase intentions towards traditional small grain foods in Zimbabwe. This model offers a novel conceptualisation of purchase intentions towards traditional small grain foods in Zimbabwe, a phenomenon that has not been comprehensively studied in the country. The model extends existing literature by providing a deeper theoretical understanding of the factors influencing consumer purchase intentions towards traditional small grain foods.

In addition, the triangulation of complementary approaches to statistical analysis offered a unique methodology never used before in consumer behaviour research on traditional foods. By employing both CB-SEM and PLS-SEM, this study leveraged complementary insights into investigated parameters thereby generating a more holistic understanding of the robustness of the model. Utilising both CB-SEM and PLS-SEM demonstrates methodological rigour and enhances the validity of the study’s findings. Finally, in a broader sense, this study addresses the identified gaps in African literature on consumer behaviour as identified by García-Barrón et al. (2021) and Moyo et al. (2023).

Implications for research

The utility of a moderately fitting structural model

The structural model showed a moderate fit to the data, as evidenced by meeting the criteria for half of the evaluated indices (50%). While the model did not pass the test for some model fit indices, it demonstrated substantial explanatory power, accounting for about 60% of the observed variance in purchase intentions. This demonstrates the practical utility of the model and emphasises that a model may still be practically useful despite not fitting the data perfectly. This reinforces Stone’s (2021) caution against focusing solely on achieving a perfect fit, which could lead to the rejection of practically useful models. The main lesson for consumer behaviour researchers is that a model’s value should be assessed based on a combination of its theoretical foundations, fit, and the practical utility it provides, as reflected by its explanatory and predictive power.

The value of triangulation in structural model analysis

The study adopted a sequential CB-PLS quantitative triangulation methodology that demonstrated utility in providing a holistic assessment of the structural model. The lesson for researchers is that even in structural equation modelling, there are opportunities for confirming the robustness of a model through triangulated analysis.

Implications for policy

Pricing and economic policies

The moderate effect size (f² = 0.035) of price illustrates the importance of price in consumer decisions. Favourable economic policies impacting traditional small grain food prices along the supply chain could be considered as they could significantly improve purchase intentions. Subsidies for traditional small grain foods or taxes on unhealthy foods could be considered by government. This could help ensure that healthier options remain affordable for consumers, thereby encouraging purchase intentions.

Public health campaigns

The small effect of ‘threat of non-communicable diseases’ on purchase intentions suggests that public health campaigns focusing solely on this threat may be insufficient to change consumer purchase intentions (f² = 0.001). More holistic campaigns aimed at improving factors with higher effect sizes, like attitude (f² = 0.103), cultural identity (f² = 0.033) and perceived behavioural control (f² = 0.011) are needed.

Implications for market researchers

Understanding consumer behaviour

The study provides valuable insights into the factors influencing consumer purchase intentions for traditional small grain foods in Zimbabwe. Market researchers can leverage these insights to develop targeted strategies that resonate with consumers. For example, the study highlights key factors such as attitudes and cultural identity that significantly influence purchasing decisions. Researchers should prioritize these elements when designing marketing campaigns and interventions. Similarly, an understanding of consumer behaviour across various consumer demographics could be useful in segmenting and targeting the market, and positioning products appropriately.

Model validation

The validated Traditional Small Grain Food Preference Model serves as a reliable framework for future market research. Practitioners can use this model to assess consumer preferences and predict purchasing behaviours in similar contexts.

Practical utility of models

The findings emphasise that a model can be practically useful even with moderate fit indices. Market researchers should focus on the practical applicability of their models rather than sticking to those with a perfect statistical fit.

Marketing budget optimization

The results of effect size (f²) estimations are useful in guiding marketing resource allocation. To optimise returns on marketing investments, more resources ought to be allocated to the most influential factor, with surplus resources being allocated to the next important factor, and so forth.

Research limitations

Prior studies have found that factors affecting food preference may vary across countries (Paul et al., 2016). This suggests that while factors influencing consumer preference have been empirically established for Zimbabwe, these may differ slightly in other contexts. This limitation implies that caution is required when applying the measurement and structural model in other African countries, particularly those with cultural differences from Zimbabwe (outside southern Africa).

Apart from the validity of the model in different contexts, the generalisation of results obtained in Zimbabwe to other African countries is contingent upon various socio-economic, cultural and contextual factors (Makanyeza & Toit, 2017). Due to shared regional characteristics between Zimbabwe and surrounding countries such as Zambia, Botswana, Namibia, South Africa and Mozambique; results obtained in Zimbabwe may be applicable in the other countries. However, significant differences exist between Zimbabwean and far afield countries such as those in West and North Africa. Extending Zimbabwean results to these countries would be discouraged prior to cross-cultural comparison studies.

The use of non-probability quota sampling in this quantitative study is acknowledged as a limitation. Within a quota, sample elements are often selected based on convenience rather than randomisation (Welch, 2014), as the case was in this study, potentially introducing bias and preventing the calculation of margins of error typical of probability sampling (Szafran et al., 2017). Despite this limitation, quota sampling was still selected for two reasons: necessity and precedence.

Regarding necessity, the absence of a sampling frame made random sampling, particularly stratified random sampling by province, practically impossible. Quota sampling emerged as the only non-probability technique comparable to random sampling. O’Connor et al. (2018) argue that a well-structured quota sampling design can be as effective as probability sampling, especially when a specific demographic variable, such as province of residence, is crucial for the study. When conducted thoughtfully and with a clear understanding of the population’s characteristics, quota sampling can facilitate the generalisation of findings to the entire population (Burns et al., 2014).

Regarding precedence, Burns and Bush (2014) assert that while probability sampling is the preferred method for achieving generalisable results, statistical inferences frequently rely on quota samples and other non-random techniques. Recent quantitative marketing studies (Lesschaeve et al., 2021; Mntande et al., 2022; Suryani et al., 2021) have employed quota sampling. Despite its limitations, quota sampling can be used cautiously in well-designed studies, allowing for a careful generalisation of results to the broader population, as this study has done.

Finally, the exclusion of individuals aged 65 or older may restrict the generalisability of findings to older adult populations. This exclusion was necessitated by ethical considerations based on institutional guidelines for ethical approval at the time of the study. Relaxing the eligibility criteria to include older participants could have potentially enriched the study by generating additional insights and perspectives unique to this demographic.

Directions for future research

The contextualisation of this study to Zimbabwe opens opportunities for the Traditional Small Grain Food Preference Model to be empirically tested in other African countries. This would allow researchers across the continent to assess the model’s robustness in different settings. Furthermore, the Traditional Small Grain Food Preference Model could be employed to develop practical marketing strategies and interventions for increasing the demand for traditional small grain foods in Zimbabwe. To achieve this, further research into how responses to measured factors differ across demographic groups is necessary, which would aid the formulation of segmentation, targeting and positioning strategies. Additionally, regression analysis could be employed to examine the relative importance of each factor influencing purchase intentions, thus enabling marketing practitioners to better allocate resources to marketing initiatives.

Furthermore, while the ability to explain 60% of the observed variability (R² = 0.595) is very significant, it shows that there is more to learn about drivers of purchase intentions towards traditional small grain foods in Zimbabwe. This is another opportunity for future research that could explore alternative model specifications to potentially enhance the model’s fit, explanatory power and predictive capability, provided such modifications are theoretically justifiable (Stone, 2021). Possible modifications could include increasing the number of factors or valid measurement items in the model, trying alternative configurations of construct relationships, or a combination of these approaches.

Conclusion

The study reveals important insights about traditional small grain food preference, building on the rich history of gastronomic practices in Zimbabwe. First, it confirms a validated model for understanding why consumers choose traditional grains like sorghum, millet and rapoko over more commonly consumed staples such as maize. This model highlights key factors that influence purchase intentions, including consumer attitudes, cultural identity and perceived control over their choices. This indicates that marketing strategies should focus on enhancing positive consumer attitudes and cultural connections to traditional grains.

The study also emphasizes the importance of pricing. It suggests that favorable economic policies, such as subsidies for traditional grains, could significantly boost their appeal. Moreover, the findings point out that public health campaigns should take a broader approach. Focusing solely on health risks may not be enough to change consumer habits. Instead, campaigns should also highlight the positive aspects of traditional grains, such as their taste and cultural relevance.

Ultimately, this research contributes valuable knowledge to the ongoing conversation about traditional small grain foods in Zimbabwe. It underscores the potential for these grains to regain their place in the diet of Zimbabweans, supported by informed marketing and policy strategies. This renewed focus could help connect consumers with their cultural heritage while also addressing modern health and environmental challenges.

Supplemental Material

Supplemental Material - Modelling purchase intentions towards traditional small grain foods in Zimbabwe

Supplemental Material for Modelling purchase intentions towards traditional small grain foods in Zimbabwe by Arnold Moyo, Felix Amoah and Marlé van Eyk in International Journal of Market Research.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical statement

ORCID iD

Arnold Moyo

Data Availability Statement

The datasets generated during and/or analysed during the current study are available from the corresponding author on request.*

Supplemental Material

Supplemental material for this article is available online.

References

Afolabi

E. O.

Ogidi

C. O.

Akinyele

B. J.

(2020). First report of nutritional value and consumer acceptability of ‘Kati’ produced from sorghum using lactic acid bacteria as starter cultures. Carpathian Journal of Food Science and Technology, 12(3), 156–166. https://doi.org/10.34302/CRPJFST/2020.12.3.13

Ahmad

Idris

Mason

Chow

(2019). The impact of young celebrity endorsements in social media advertisements and brand image towards the purchase intention of young consumers. International Journal of Financial Research, 10(5), 54–65. https://doi.org/10.5430/ijfr.v10n5p54

Ajzen

(1991). The theory of planned behaviour. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T

Ajzen

(2010). Constructing a theory of planned behaviour questionnaire. Biofeedback and Selfregulation, 17(March), 1–7.

Ajzen

, (2020). The theory of planned behaviour: Frequently asked questions. 2(April), 314-324. https://doi.org/10.1002/hbe2.195

Akinola

Pereira

L. M.

Mabhaudhi

de Bruin

F. M.

Rusch

(2020). A review of indigenous food crops in Africa and the implications for more sustainable and healthy food systems. Sustainability, 12(8), 1–30. https://doi.org/10.3390/su12083493

Albarracín

Johnson

B. T.

Fishbein

Muellerleile

P. A.

(2016). Theories of reasoned action and planned behaviour as models of condom use: A meta-analysis. Psychological Bulletin, 127(1), 142–161. https://doi.org/10.1037/0033-2909.127.1.142

Anderson

J. C.

Gerbing

D. W.

(1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103(3), 411–423. https://doi.org/10.1037/0033-2909.103.3.411

Armitage

C. J.

Conner

Norman

(1999). Differential effects of mood on information processing: Evidence from the theories of reasoned action and planned behaviour. European Journal of Social Psychology, 29(4), 419–433. https://doi.org/10.1002/(sici)1099-0992(199906)29:4<419::aid-ejsp933>3.0.co;2-l

10.

Banerjee

Maitra

(2020). The role of small millets as functional food to combat malnutrition in developing countries. Indian Journal of Natural Sciences, 10(60), 20412–20417.

11.

Bechoff

Chijioke

Westby

Tomlins

K. I.

(2018). Yellow is good for you. Consumer perception and acceptability of fortified and biofortified cassava products. PLoS One, 13(9), 1–22. https://doi.org/10.1371/journal.pone.0203421

12.

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychological bulletin, 107(2), 238–246. https://doi.org/10.1037/0033-2909.107.2.238

13.

Blythe

(2013). Consumer behaviour (2nd ed.), Sage.

14.

Chauke

D. X.

Duh

H. I.

(2019). Marketing and socio-psychological factors influencing organic food purchase and post-purchase outcomes. Journal of Food Products Marketing, 25(9), 896–920. https://doi.org/10.1080/10454446.2019.1697980

15.

Burns

A. C.

Bush

R. F.

Sinha

(2014). Marketing research (8th ed.). Boston: Pearson.

16.

Chauke

X. D.

(2018). Market and socio-psychological factors affecting organic food purchase decision and post-purchase outcomes in South Africa. University of the Witwatersrand.

17.

Chen

(2012). A study investigating the determinants of consumer buyer behaviour relating to the purchase of organic food products in urban China Unpublished PhD Thesis. Swinburne University of Technology.

18.

Chen

Deng

(2016). Research on the green purchase intentions from the perspective of product knowledge. Sustainability, 8(9), 1–16. https://doi.org/10.3390/su8090943

19.

Chen

P.-J.

Antonelli

(2020). Conceptual models of food choice: Influential factors related to foods, individual differences, and society. Foods, 9(12), 1898. https://doi.org/10.3390/foods9121898

20.

Choi

Park

Jeon

Asperin

A. E.

(2021). Exploring local food consumption in restaurants through the lens of locavorism. Journal of Hospitality Marketing & Management, 30(8), 982–1004. https://doi.org/10.1080/19368623.2021.1923608

21.

Chopera

Zimunya

Mugariri

Matsungo

(2022). Facilitators and barriers to the consumption of traditional foods among adults in Zimbabwe. Journal of Ethnic Foods, 9(1), 5. https://doi.org/10.1186/s42779-022-00121-y

22.

Coley

K. M.

Perosky

J. E.

Nyanplu

Kofa

Anankware

J. P.

Moyer

C. A.

Lori

J. R.

(2020). Acceptability and feasibility of insect consumption among pregnant women in Liberia. Maternal and Child Nutrition, 16(3), e12990. https://doi.org/10.1111/mcn.12990

23.

Conner

Armitage

C. J.

(2002). The social psychology of food. Open University Press.

24.

Dash

Paul

(2021). CB-SEM vs PLS-SEM methods for research in social sciences and technology forecasting. Technological Forecasting and Social Change, 173(December), Article 121092. https://doi.org/10.1016/j.techfore.2021.121092

25.

de Beer

Mielmann

Coetzee

(2016). Exploring the acceptability of amaranth-enriched bread to support household food security. British Food Journal, 118(11), 2632–2646. https://doi.org/10.1108/BFJ-06-2016-0240

26.

Doan

H. Q.

(2021). Critical factors affecting consumer buying behaviour of organic vegetables in Vietnam. Journal of Asian Finance, 8(9), 333–340. https://doi.org/10.13106/jafeb.2021.vol8.no9.0333

27.

Ercis

Yildiz

Deveci

F. G.

(2020). Investigation of consumers´ organic food purchases in the context of the relationship of personal values and individual factors. Journal of Mehmet Akif Ersoy University Economics and Administrative Sciences Faculty, 7(2), 297–325. https://doi.org/10.30798/makuiibf.573331

28.

Fang

(2022). Research on the influence of interaction with the streamer on customers’ purchase intention in the context of live-streaming e-commerce. Social Sciences, 11(6), 373–380. https://doi.org/10.11648/j.ss.20221106.14

29.

Fernández

J. G.

Martelo-Landroguez

Vélez-Colón

Cepeda‐Carrión

(2018). An explanatory and predictive PLS-SEM approach to the relationship between organizational culture, organizational performance and customer loyalty. Journal of Hospitality and Tourism Technology, 9(3), 438–454. https://doi.org/10.1108/jhtt-09-2017-0100

30.

Ferreira

Pereira

(2023). Antecedents of consumers’ intention and behavior to purchase organic food in the Portuguese context. Sustainability, 15(12), 1–13. https://doi.org/10.3390/su15129670

31.

Frez-Muñoz

Kampen

J. K.

Fogliano

Steenbekkers

B. L. P. A.

(2021). The food identity of countries differs between younger and older generations: A cross-sectional study in American, European and asian countries. Frontiers in Nutrition, 8, Article 653039. https://doi.org/10.3389/fnut.2021.653039

32.

Garbacz

(2017). Consumers’ attitudes and behaviour towards local food in Poland. In: International master of science in rural development: Ghent University. https://lib.ugent.be/catalog/rug01:002482265

33.

García-Barrón

S. E.

Leyva-Trinidad

D. A.

Carmona-Escutia

R. P.

Romero-Medina

Villanueva-Rodríguez

S. J.

(2021). Evaluation of traditional foods from a consumer’s perspective: A review of methods and highlights. British Food Journal, 123(12), 4451–4465.

34.

Gorton

Barjolle

(2013). Theories of food choice. In Barjolle

Dordević

J. M.

Gorton

Stojanović

(Eds.), Food consumer science: Theories, methods and application to the Western Balkans (pp. 15–26): Springer Science.

35.

Gupta

Qumar

Noor

(2023). Awareness and consumption of millets among women. International Journal of Horticulture and Food Science, 5(1), 46–49. https://doi.org/10.33545/26631067.2023.v5.i1a.156

36.

Hair

J. F.

Black

W. C.

Babin

B. J.

Anderson

R. E.

Tatham

R. L.

(2014). Multivariate data analysis (7th ed.), Prentice Hall.

37.

Hair

J. F.

Hult

G. T. M.

Ringle

C. M.

Sarstedt

Danks

N. P.

Ray

(2021). Partial least squares structural equation modeling (PLS-SEM) using R: A workbook. Springer.

38.

Hansen

Sørensen

M. I.

Eriksen

M. L. R.

(2018). How the interplay between consumer motivations and values influences organic food identity and behaviour. Food Policy, 74(January), 39–52.

39.

Hassan

Z.M.

Sebola

N.A.

Mabelebele

(2021). The nutritional use of millet grain for food and feed: A review. Agriculture & Food Security, 10(1), 1–14. https://doi.org/10.1186/s40066-020-00282-6

40.

Hawkins

D. I.

Mothersbaugh

D. L.

(2013). Consumer behaviour: Building marketing strategy (12th ed.), McGraw-Hill.

41.

Hong

Yao

Xie

(2020). An empirical study on consumer purchase intention for nuts and influencing factors: Survey based on consumers from Zhejiang. Food Control, 117(November), Article 107343. https://doi.org/10.1016/j.foodcont.2020.107343

42.

Hsu

(2014). Food tourism: Consumer behaviour in relation to traditional food. Doctoral Thesis. University of Queensland.

43.

L. T.

Bentler

P. M.

(1995). Evaluating model fit. In Hoyle

R. H.

(Ed.), Structural equation modeling: Concepts, issues and application (1st ed., pp. 77–99). California: Sage.

44.

Ismael

Duleba

(2021). Investigation of the relationship between the perceived public transport service quality and satisfaction: A PLS-SEM technique. Sustainability, 13(23), 1–20. https://doi.org/10.3390/su132313018

45.

Jane

M. M.

Lodewyk

Elma

Carlien

Zvauya

(2015). Characterisation of yeasts isolated from traditional opaque beer beverages brewed in zimbabwean households. African Journal of Microbiology Research, 9(8), 549–556. https://doi.org/10.5897/ajmr2014.7218

46.

Jeong

J. Y.

Kim

H. C.

(2020). Korean mothers’ food choice behavioural intent for children: An examination of the interaction effects of food type, household income, and healthism. Food Quality and Preference, 81(April), Article 103835. https://doi.org/10.1016/j.foodqual.2019.103835

47.

Joshi

Uniyal

D. P.

Sangroya

(2021). Investigating consumers’ green purchase intention: Examining the role of economic value, emotional value and perceived marketplace influence. Journal of Cleaner Production, 328(15 December), Article 129638. https://doi.org/10.1016/j.jclepro.2021.129638

48.

Kauma

Swart

(2022). Hunger and power: Politics, food (in)security and the development of small grains in Zimbabwe, 2000-2010. Historia, 67(1), 1–33. https://doi.org/10.17159/2309-8392/2021/v67n1a6

49.

Kim

J. Y.

Park

Y. J.

Park

Y. B.

(2016). Confirmatory and Exploratory Factor Analysis for validating the phlegm pattern questionnaire for healthy subjects: Evidence-based complementary and alternative medicine. eCAM. https://doi.org/10.1155/2016/2696019

50.

Kudita

Schoustra

Mubaiwa

Smid

Linnemann

(2023). Substitution of maize with sorghum and millets in traditional processing of mahewu, a non‐alcoholic fermented cereal beverage. International Journal of Food Science and Technology, 59(3), 1421–1431. https://doi.org/10.1111/ijfs.16887

51.

Kung

M. L.

Wang

J. H.

Liang

(2021). Impact of purchase preference, perceived value, and marketing mix on purchase intention and willingness to pay for pork. Foods, 10(10), 1–17. https://doi.org/10.3390/foods10102396

52.

Lesschaeve

Glaurdić

Mochťak

(2021). Health versus wealth during the covid-19 pandemic. Public Opinion Quarterly, 85(3), 808–835. https://doi.org/10.1093/poq/nfab036

53.

Loh

Hassan

S.H.

(2022). Consumers’ attitudes, perceived risks and perceived benefits towards repurchase intention of food truck products. British Food Journal, 124(4), 1314–1332. https://doi.org/10.1108/BFJ-03-2021-0216

54.

López-Parra

Zamora-Carmona

Sianes-Gallén

López-González

Gil-Rey

Costa-Ventura

Borrás-Sánchez

Rayo-Posadas

Arizu-Puigvert

Vives-Vilagut1

(2022). Patient information and informed consent for research in elderly: Lessons learned from a randomized controlled trial. Research Square, 1–13. https://assets.researchsquare.com/files/rs-1169657/v1/e8708b35-c24d-4216-ac28-5946b78684d7.pdf?c=1649110830 (Accessed 10 July 2021).

55.

Makanyeza

Toit

F. D.

(2017). Consumer ethnocentrism in developing countries: Application of a model in Zimbabwe. Acta Commercii, 17(1), 481. https://doi.org/10.4102/ac.v17i1.481

56.

Maumbe

B. M.

Chikoko

(2020). The rise of supermarkets in Zimbabwe against a tide of macroeconomic uncertainty. Development Southern Africa, 1(1), 1–14. https://doi.org/10.1080/0376835x.2020.1832443

57.

McCartan

Van Burgel

McArthur

Testa

Thurn

Funston

Kho

McMahon

Brimblecombe

(2020). Traditional food energy intake among indigenous populations in select high-income settler-colonized countries: A systematic literature review. Current Developments in Nutrition, 4(11), 1–26. https://doi.org/10.1093/cdn/nzaa163

58.

Mntande

K. A.

Stiehler-Mulder

Roberts-Lombard

(2022). Securing delight and loyalty in a market with low switching costs. European Business Review, 35(1), 1–22. https://doi.org/10.1108/ebr-03-2022-0043

59.

Moyo

Amoah

van Eyk

(2021). A proposed hypothetical framework for investigating antecedents of consumer purchase intentions towards traditional foods. 22nd PARIS Int'l Conference on Marketing, Education, Social Sciences & Humanities (MESSH-21). Paris (France), Dec. 730-935. Parish: Excellence in Research & Innovation. https://doi.org/10.17758/EIRAI11.EAP1221410

60.

Moyo

Amoah

van Eyk

(2023). Consumer behaviour research on traditional foods in Africa: A scoping review. Cogent Business & Management, 10(2), 1–16. https://doi.org/10.1080/23311975.2023.2213532

61.

Mushore

Mhizha

Manjowe

Mashawi

Matandirotya

Mashonjowa

Mushambi

Gwenzi

(2021). Climate change adaptation and mitigation strategies for small holder farmers: A case of nyanga district in Zimbabwe. Frontiers in Climate, 3, 676495. https://doi.org/10.3389/fclim.2021.676495

62.

Muthamilarasan

Prasad

(2021). Small millets for enduring food security amidst pandemics. Trends in Plant Science, 26(1), 33–40. https://doi.org/10.1016/j.tplants.2020.08.008

63.

Muvhuringi

P. N.

Terrence

Chigede

(2021). The impact of COVID-19 on agricultural extension and food supply in Zimbabwe. Cogent Food & Agriculture, 7(1), Article 1918428. https://doi.org/10.1080/23311932.2021.1918428

64.

Ndlovu

Prinsloo

Roux

(2020). Impact of climate change and variability on traditional farming systems: Farmers’ perceptions from south-west, semi-arid Zimbabwe. Jàmbá Journal of Disaster Risk Studies, 12(1), 742. https://doi.org/10.4102/jamba.v12i1.742

65.

Novitasari

Chidir

Sutardi

Iskandar

Pebrina

(2021). The role of Halal awareness on purchase intention of Halal food: Evidence from Indonesian working students. International Journal of Science and Management Studies, 4(1), 31–44. https://doi.org/10.51386/25815946/ijsms-v4i1p104

66.

Nyasha

Tonderai

Tinyiko

(2023). Opportunities, perceptions and challenges in production and use of traditional grains in semi-arid masvingo district, Zimbabwe. African Journal of Agricultural Research, 19(8), 849–856. https://doi.org/10.5897/ajar2023.16439

67.

Ockey

G. J.

Choi

(2015). Structural equation modeling reporting practices for language assessment. Language assessment quarterly, 12, 305–319. https://doi.org/10.1080/15434303.2015.1050101

68.

O’Connor

R. C.

Wetherall

Cleare

Eschle

Drummond

Ferguson

O'Carroll

R. E.

O’Carroll

R. E.

(2018). Suicide attempts and non-suicidal self-harm: National prevalence study of young adults. BJPsych Open, 4(3), 142–148. https://doi.org/10.1192/bjo.2018.14

69.

Paul

Modi

Patel

(2016). Predicting green product consumption using theory of planned behaviour and reasoned action. Journal of Retailing and Consumer Services, 29(March), 123–134. https://doi.org/10.1016/j.jretconser.2015.11.006

70.

Penhill

(2020). What is the difference between satiety and satiation quizlet? Available: https://short-fact.com/what-is-the-difference-between-satiety-and-satiation-quizlet/ (Accessed 15 August 2022).

71.

Phiri

Dube

Moyo

Ncube

Ndlovu

(2019). Small grains “resistance”? Making sense of Zimbabwean smallholder farmers’ cropping choices and patterns within a climate change context. Cogent Social Sciences, 5(1), 1–13. https://doi.org/10.1080/23311886.2019.1622485

72.

Pramana

A. I. M.

Fachrunnisa

(2024). Model of Improving human resource performance through achievement orientation, skill level and individual factors. Journal of Management and Science, 14(1), 77–86. https://doi.org/10.26524/jms.14.10

73.

Putra

M. F. P.

Guntoro

T. S.

Wandik

Ita

Sinaga

Hidayat

R. R.

Singa

Muhammad

Singa

F. S. G.

Kmurawak

R. M. B.

Rahayu

A. S.

(2021). Psychometric Properties at Indonesian Version of the Sport Anxiety Scale-2: Testing on Elite Athletes of Papua, Indonesia. International Journal of Human Movement and Sports Sciences, 9(6), 1477–1485.

74.

Rampa

Lammers

Linnemann

Schoustra

De Winter

(2020). Pathways to improved food and nutrition security of the poor: The promise of African indigenous foods and technologies. https://knowledge4food.net/wp-content/uploads/2020/03/FBRsynthesis02_LocalIndigenousFoods_paper.pdf (Accessed 2 March 2022).

75.

Randall

Sanjur

(1981). Food preferences: Their conceptualization and relationship to consumption. Ecology of Food and Nutrition, 11(3), 151–161. https://doi.org/10.1080/03670244.1981.9990671

76.

Review World Population . (2022). Zimbabwe population 2022 (Demographics, Maps, Graphs). https://worldpopulationreview.com/countries/zimbabwe-population (Accessed 15 December 2023).

77.

Saleh

Zhang

Chen

Shen

(2013). Millet grains: Nutritional quality, processing, and potential health benefits. Comprehensive Reviews in Food Science and Food Safety, 12(3), 281–295. https://doi.org/10.1111/1541-4337.12012

78.

Saunders

M. N. K.

Lewis

Thornhill

(2019). Research methods for business students (8th ed.), Pearson Education Limited.

79.

Senevirathne

I. G. N. H.

Abeysekera

W. K. S. M.

Abeysekera

W. P. K. M.

Jayanath

N. Y.

Galbada Arachchige

S. P.

Wijewardana

D. C. M. S. I.

(2021). Antiamylase, antiglucosidase, and antiglycation properties of millets and sorghum from Sri Lanka. Evidence-based Complementary and Alternative Medicine: eCAM, 2021(18 June), 1–10. https://doi.org/10.1155/2021/5834915

80.

Sheethal

Baruah

Kalpuri

Ananthan

Longvah

(2022). Insights of nutritional and anti-nutritional retention in traditionally processed millets. Frontiers in Sustainable Food Systems, 5, 735356. https://doi.org/10.3389/fsufs.2021.735356

81.

Shen

Cao

Esfahani

S. S.

Saleem

(2022). Factors influencing consumers’ purchase intention on cold chain aquatic products under COVID-19: An investigation in China. International Journal of Environmental Research and Public Health, 19(8), 4903. https://doi.org/10.3390/ijerph19084903

82.

Shen

Y.-C.

Chen

H.-S.

(2020). Exploring consumers’ purchase intention of an innovation of the agri-food industry: A case of artificial meat. Foods, 9(6), 1–15. https://doi.org/10.3390/foods9060745

83.

Shmueli

Sarstedt

Hair

J. F.

Cheah

Ting

Vaithilingam

Ringle

C. M.

(2019). Predictive Model Assessment in PLS-SEM: Guidelines for using PLSpredict. European Journal of Marketing, 53(11), 2322–2347.

84.

Sinha

Annamdevula

(2022). The antecedents of green purchase behaviour of Indian households. Electronic Green Journal, 1(47), 1–16. https://doi.org/10.5070/g314752650

85.

Soper

D. S.

(2020). A-priori sample size calculator for structural equation models. [Software]. https://wwwdanielsopercom/statcalc

86.

Steenkamp

J.-B. E. M.

(1997). Dynamics in consumer behaviour with respect to agricultural and food products. In Wierenga

van Tilburg

Grunert

K. G.

Steenkamp

J.-B. E. M.

Wedel

(Eds.), Agricultural marketing and consumer behaviour in a changing world (pp. 143–188): Springer.

87.

Stone

B. M.

(2021). The ethical use of fit indices in structural equation modeling: Recommendations for psychologists. Frontiers in Psychology, 12, 1–4. https://doi.org/10.3389/fpsyg.2021.783226

88.

Suryani

Abdinagoro

S. B.

Arief

M. R.

Hamsal

(2021). Impact of digital literacy and market orientation through e-commerce adoption on the msme performance moderated by o2o business adoption. Proceeding of the International Conference on Family Business and Entrepreneurship, 2(1), 3544. https://doi.org/10.33021/icfbe.v2i1.3544

89.

Szafran

R. F.

Cordova

Forbes

Guidry

Montoya

(2017). Restoring order. Journal of Applied Social Science, 11(1), 60–69. https://doi.org/10.1177/1936724417692375

90.

Tanyanyiwa

(2021). Hunger by choice? Rethinking food security strategies. European Journal of Development Studies, 1(2), 17–21. https://doi.org/10.24018/ejdevelop.2021.1.2.17

91.

Thilina

Gunawardane

(2019). The effect of perceived risk on the purchase intention of electric vehicles: An extension to the technology acceptance model. International Journal of Electric and Hybrid Vehicles, 11(1), 73–84. https://doi.org/10.1504/ijehv.2019.10020148

92.

Verbeke

(2005). Consumer acceptance of functional foods: Socio-demographic, cognitive and attitudinal determinants. Food Quality and Preference, 16(1), 45–57. https://doi.org/10.1016/j.foodqual.2004.01.001

93.

Welch

(2014). American opinion toward jews during the nazi era: Results from quota sample polling during the 1930s and 1940s. Social Science Quarterly, 95(3), 615–635. https://doi.org/10.1111/ssqu.12084

94.

Williams

S. C. P.

(2021). What Is physiology? Available at: https://www.webmd.com/a-to-z-guides/what-is-physiology (Accessed 20 August 2022).

95.

Zayed

M. F.

Gaber

H. R.

el Essawi

(2022). Examining the factors that affect consumers’ purchase intention of organic food products in a developing country. Sustainability, 14(10), 1–16.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.28 MB