Table 2 presents the demographic characteristics of the respondents as well as their frequency of consumption of products from multinational food corporations. The study comprised 200 participants with a median age of 26.5 years (interquartile range (IQR): 22–32 years). Most of the respondents were female (53%), unmarried (70.5%), childless (66.5%), and held a bachelor's degree (38%). These findings mirror the expected composition of the study population, as the Greater Accra region is known to have a distinct age distribution, with a higher proportion of young adults (aged 15–35 years) and a lower total fertility rate (2.2%) compared to other regions in Ghana [34]. Among the respondents, 38.5% reported frequently consuming products or meals from multinational food corporations or fast-food chains during the daytime, with 32.5% reporting occasional consumption, 16% reporting habitual consumption, 12.5% reporting infrequent consumption, and only 0.5% reporting never consuming such products. Respondents aged 15–29 years, females, unmarried individuals, childless individuals, and those with a bachelor's degree reported higher rates of frequent consumption, with 25.5%, 21%, 29.5%, 28%, and 16% of respondents in these respective categories reporting frequent consumption of multinational food corporation products or meals.

The results of the maximum difference model (Likelihood Ratio (LR) test statistic = 1203.665, p < 0.0001) demonstrated significant differences in preferences for the various attributes (Table 3). The estimated preference coefficients exhibited the expected sign within the 95% CIs. Each attribute was statistically significant (i.e., 95% CIs did not contain zero or were greater or less than zero); however, there was considerable overlap in the 95% CIs, indicating that certain attributes were not statistically different from each other. We interpret these results with respect to the model specification, beginning with the observation that the reference level was the attribute variable visual aesthetic. Preference coefficients for the attribute variables, such as aroma/smell, availability, food quality/packaging, healthiness, image/desirability, less preparation time of a meal, nutrition content/nutritional value, social (family/friends eat), taste/flavour, and texture, were mostly significant, indicating their influence on consumers' choice decisions. Specifically, the positive signs of the preference coefficients for attributes such as aroma/smell, food quality/packaging, healthiness, nutrition content/nutritional value, and taste/flavour indicated that consumers were more likely to choose products from multinational corporations over local foods. Conversely, negative signs for attributes such as availability, image/desirability, less preparation time of a meal, social (family/friends eat), and texture indicated disutility for choosing multinational corporations' products over local foods. This suggests that consumers tended to weigh the attributes of availability, image/desirability, less preparation time of a meal, social (family/friends eat), and texture against each other when selecting multinational food corporations' products/meals.

To facilitate interpretation of the relative importance of each plausible attribute to consumers based on the magnitude of the preference coefficients, we have arranged the attributes in Fig. 2 in accordance with the marginal utility estimates (MUE) and their corresponding marginal probability (MP) values.

Marginal utility estimates and marginal probability of attributes that contribute to consumers choice of multinational corporations’ food products over local meals

MUE represents the perceived importance of the matching level of the effect. Larger values imply that the level of the effect is of greater importance. MP represents the estimated probability that a consumer expresses a preference for the matching effect over all other effects. Our analysis revealed that food quality/packaging was the most important attribute (MUE: 0.8053; MP: 0.1286; 95% confidence interval [CI]: 0.7134, 0.8983), followed by healthiness (MUE: 0.5996; MP: 0.1047; 95% CI: 0.5102, 0.6896), taste/flavour (MUE: 0.4918; MP: 0.0940; 95% CI: 0.4037, 0.5805), nutrition content/nutritional value (MUE: 0.4077; MP: 0.0864; 95% CI: 0.3201, 0.4957), and aroma/smell (MUE: 0.1090; MP: 0.0641; 95% CI: 0.0226, 0.1954). However, we also observed negative relative importance for some attributes, such as availability (MUE: -0.187; MP: 0.0477; 95% CI: -0.2730, -0.1003), less preparation time of a meal (MUE: -0.233; MP: 0.0455; 95% CI: -0.3196, -0.1463), social (family/friends eat) (MUE: -0.273; MP: 0.0437; 95% CI: -0.3602, -0.1871), and image/desirability (MUE: -0.328; MP: 0.0414; 95% CI: -0.4157, -0.2416). Finally, texture was found to be the least important attribute (MUE: -0.826; MP: 0.0252; 95% CI: -0.9184, -0.7337).

Additionally, Table 4 provides a comparison of the greatest utility difference (GUD) among the preference weights for the attributes. GUD is defined as the maximum change in utility that can be achieved from an attribute, based on the plausible attributes included in the maximum difference experiment. Our results show that food quality/packaging had significantly the highest utility (GUD: 1.63088; p = 5e-113) compared to all other attributes related to consumers' preferences for multinational food corporations’ products/meals over local foods. Healthiness also had a significantly higher utility (GUD: 1.42512; p = 4.1e-91) but did not differ significantly from the taste/flavour attribute (p = 0.10109). The taste/flavour attribute had the next highest utility (GUD: 1.31736; p = 8.3e-80), followed by the nutrition content/nutritional value attribute (GUD: 1.23321; p = 3.4e-71), though it did not differ significantly from the taste/flavour attribute (p = 0.1957). The aroma/smell attribute had a lower but still significant utility (GUD: 0.93449; p = 1.7e-43), though it did not differ significantly from the convenience attribute (p = 0.10964).

In our study, we found that the affordability attribute had a positive utility value, indicating that it is a desirable attribute for consumers (GUD: 0.88151; p = 4.3e-39). However, this attribute did not significantly differ from other attributes such as high fibre and roughage, familiarity of a meal, convenience, and aroma/smell. When affordability was made available concurrently with other attributes, such as food quality/packaging, healthiness, nutrition content/nutritional value, and taste/flavour, consumers tended to trade it off (GUDs ranging from -0.3517 to -0.7494; ps < 0.05), suggesting that affordability may not be the deciding factor in their meal choices. Similarly, the convenience attribute had a significant positive utility value (GUD: 0.83257; p = 1.6e-35), but it did not differ significantly from the high fibre and roughage and familiarity of a meal attributes. When convenience was made available concurrently with other attributes such as food quality/packaging, healthiness, nutrition content/nutritional value, and taste/flavour, consumers tended to trade it off (GUDs ranging from -0.4016 to -0.7993; ps < 0.05). Overall, our findings suggest that consumers prioritize attributes such as food quality/packaging, healthiness, nutrition content/nutritional value, and taste/flavour over affordability and convenience when selecting international meal or food products over local options.

We also observed significant differences for the attribute variable high in fibre and roughage (DUE: 0.79409; p = 2.2e-32). However, when made available concurrently with the nutrition content/nutritional value attribute (GUD: -0.4391; p = 1.4e-11) and the taste/flavour attribute (GUD: -0.5233; p = 19.9e-16), it will be traded off. Similarly, the accessibility attribute was significantly different (GUD: 0.79298; p = 3.9e-32), but it will be traded off when made available concurrently with the aroma/smell attribute (GUD: -0.1415; p = 0.02891), healthiness (GUD: -0.6321; p = 9.2e-22), food quality/packaging (GUD: -0.8379; p = 1.4e-35), nutrition content/nutritional value (GUD: -0.4402; p = 1.5e-11) as well as taste/flavour (GUD: -0.5244; p = 31.1e-15). These findings suggest that the high in fibre and roughage attribute and the accessibility attribute are not as important as the other attributes when making food choices.

The study found that the attribute variable "familiarity of a meal" had a significant positive utility value (GUD: 0.76385; p = 6.6e-30), but it did not differ significantly from the attribute "high in fibre and roughage" (p = 0.63583) or "availability" (GUD: 0.63892; p = 1.2e-21), which did not differ significantly from the attribute "less preparation time of a meal" (p = 0.47014) or "social (family/friends eat)" (p = 0.17776). However, when made available concurrently with other attributes, familiarity of a meal was traded-off for food quality/packaging (GUD: -0.867; p = 5.3e-38), healthiness (GUD: -0.6613; p = 1.3e-23), nutrition content/nutritional value (GUD: -0.4694; p = 5.1e-13), and taste/flavour (GUD: -0.5535; p = 1.8e-17). Similarly, "less preparation time of a meal" had a significant utility value (GUD: 0.59271; p = 8.6e-19), but it did not differ significantly from the "social (family/friends eat)" attribute (p = 0.52704). However, when made available concurrently with other attributes, "less preparation time of a meal" was traded-off for nutrition content/nutritional value (GUD: -0.6405; p = 1.6e-22) and taste/flavour (GUD: -0.7246; p = 34.2e-28). Finally, the attribute variable "social (family/friends eat)" had a significant positive utility value (GUD: 0.55203; p = 91.2e-16), but it was traded-off for taste/flavour (GUD: -0.7653; p = 63.7e-31) when made available concurrently with this attribute.

The attribute variable "image/desirability" was found to have a significantly lower utility (GUD: 0.49703; p = 8.7e-14) than all other attributes related to multinational food corporations’ products/meals preference over local foods. However, this attribute did not differ significantly from the attributes of "less preparation time of a meal" (p = 0.13785) and "social (family/friends eat)" (p = 0.39453). These results suggest that although "image/desirability" can influence consumers' food choices, it is less valued than other attributes. It is noteworthy that the "image/desirability" attribute will be traded off when made available concurrently with "nutrition content/nutritional value" (GUD: -0.7362; p = 5.6e-29) and "taste/flavour" (GUD: -0.8203; p = 3.1e-35). Thus, consumers are likely to base their food choices on nutrition content/nutritional value and taste/flavour rather than the image/desirability of the meals/products.