1. Introduction
The Food and Agriculture Organization (FAO) forecasts a 2.7% annual increase in meat production in the coming years, primarily driven by population growth in developing countries [
1]. It is widely accepted that by 2050 there will be nine billion people in the world [
1,
2,
3]. To cope with this number, current food production will need to almost double. This increase, along with the limited land area available, may pose a challenge for the meat industry worldwide, and alternative protein sources will most likely be needed to feed a growing population. We need to find new ways of growing food [
2,
3,
4]. Current production techniques, in particular animal breeding for protein production, cause greenhouse gas emissions (GHGs), require high water inputs and generate considerable amounts of waste. Insect production is generally considered to be more sustainable from an environmental point of view [
5,
6]. The deficit in the production of protein has become one of the most important problems in Europe, as underlined by the European Parliament; in fact, about 80% of European protein crop requirements are imported from non-EU countries [
7,
8]. In particular, the development and exploitation of new sources of sustainable food ingredients is one of the new frontiers in the EU when it comes to radical innovation [
9,
10].
Meat production and consumption are considered one of the most relevant sources of environmental degradation as well as health costs related to diabetes, cardiovascular disease and obesity, mainly due to over-consumption [
9,
11,
12,
13].
Considering these premises, insects can become a sustainable production for several reasons such as: (a) high conversion ratios: the production of 1 kg of insects requires only 2 kg of food [
2,
14]; (b) breeding produces fewer pollutants and GHGs [
15]; (c) possible biological breeding [
16]; (d) reduced use of soil and water compared to livestock farms [
17]; and (e) less transmissibility to humans and animals of pathogens. The latter point however needs to be more carefully verified. Entomophagy, or eating insects, is a very common practice in Asia, Africa, and Latin America. It is estimated that insects form part of the traditional diets of at least two billion people. More than 2100 species have reportedly been used as food [
18].
In Western countries, although several studies demonstrate the benefits of using insect flours [
19,
20,
21], arthropod consumption is not part of traditional eating habits and very often causes reactions of disgust. Indeed, accepting or rejecting the food is linked to various reasons: (a) the characteristics of the food; (b) cultural motivations; (c) individual preferences. Insects are an unfamiliar food source for most European consumers, and sources that deviate from cultural norms can attract consumers who constantly sample new and unusual foods while potentially alienating others who are less adventurous [
22,
23,
24,
25]. The latter tendency is commonly denoted as food neophobia, a personal characteristic that accounts for the wide variability among consumers in terms of their attitudes towards novel foods [
26,
27,
28]. Several studies have been conducted to understand the reasons for insect consumption (or their derivatives), and what can the factors may be that can lead to a change in eating habits [
29,
30,
31,
32].
Among the key aspects of helping people overcome their reluctance to eat insects is the role of information [
25]. Communication of information on contents positively influences the evaluation of foods containing insects. Information on the positive effects of edible insects, sustainability and environmental perspectives increases consumers’ willingness to consume insect products [
27]. An active role for public health institutions is crucial for expanding consumption [
33].
Focusing on processed insect products is shown to be the most promising strategy to implement entomophagy, as an essential barrier to consumption is the visibility of the insects [
34], the willingness to eat insect food depends on the form in which the products are presented [
35].
Furthermore, food culture and perceived social norms have substantial explanatory roles when it comes to eating strange and perhaps disgusting foods such as insects [
36].
Lack of practice in the preparation is a serious obstacle to consumption that could be overcome by proposing recipes, preparation methods or new consumption opportunities. Curiosity and environmental benefits are the most important factors for motivating insect consumption in the future [
37].
It is interesting to note that both unprocessed and processed insect products generate more positive perceptions after tasting than expected [
24].
People who have grown up and live in a strong and widespread food culture may be less likely to try new and different products than people who live in a rapidly changing food culture [
38].
The research aims to validate a questionnaire of propensity for entomopahgy, composed of 32 survey items to measure the willingness of respondents to accept insects and/or insect-based products in their diet. The methodology applied here, for the first time in this area, is that of Rasch models [
39].
3. Results and Discussion
The questionnaire aimed to measure the propensity to eat insects (entomophagy) was administered via the CASI (Computer Assisted Self Interviewing) method. Academic staff and students of a center Italian University were invited to participate in the survey. Data collection from students is often able to well represent the point of view of a larger community [
55,
56,
57].
The sample of 506 individuals is composed by 59.5% of females, the most represented age group is 20–30 (43.9%), 55.3% of the sample are graduates, 85.4% are of Italian nationality and 91.3% are omnivorous. (
Table 2).
In response to the item “How often do you eat an insect-based meal? (D9), 82.8% of the sample declared never, 10.5% consumed insects once or twice at most and the remainder (7%) consumed insects more or less regularly. Considering the two countries most represented in the survey (Italy and Mexico) it is found that 91% of Italians have never consumed insects, while this happens only for 20% of Mexicans. This clearly shows that entomophagy has a fundamentally cultural basis. In response to the item “In case of hunger or need, would you eat insects?” (D12) 7.3% of the sample declared “absolutely no”, while 70.8% of the sample were possibilist in this regard. In response to the item “Which insect(s) would you consider eating among the ones listed below?” (D22) the highest percentage (39.3%) was observed for cricket followed by grasshopper (36.6%), last in the ranking (5.1%) beetle. In response to the item “How much would you be willing to pay for 10 g of ready-to-eat insects?” (D15), 41.9% of the sample would paid nothing, but the rest (58.1%) would paid the equivalent of a hamburger or almost.
These are higher percentages than people who had real insect tasting experiences. This demonstrates that a communication process is necessary that aims to overcome psychological/cultural barriers [
72]. Only in this way will it be possible to increase the propensity to consume insects.
Most of the items had the task of assessing consumer behavior, from complete disagreement to complete agreement, with respect to entomophagy issues. For that purpose, each response was coded by the Likert scale (0, 1, 2 …) so that the smallest value corresponded to low levels of propensity for entomophagy and the highest value corresponded to the maximum propensity for entomophagy. Items that did not fit the model were then deleted. Subsequently, people who did not fit the model and who evidently tended to answer the questionnaire at random or inconsistently with the scale were also eliminated.
Table 3 shows the reliability indices for items and people.
The items selected for analysis are therefore 23, all not extreme and with a reliability index of 0.99 (
Table 3). The average measurement level of the items is conventionally set to zero. The range of Infit indexes of the items vary between 0.64 and 1.49, and the outfit indexes vary between 0.27 and 1.41, within the limits provided for the good adaptation of a Rating Scale model [
70]. The indexes, for items and people, have very high values that highlight the goodness of the scale used. Cronbach’s Alfa is also quite high and equal to 0.90. The average level of all non-extreme and extreme people (the latter answered all items of the questionnaire with minimum or maximum values on the scale), is −2.04, which is lower than the previous one and indicates that most extreme people answered all items with the lowest mode on the scale therefore very low propensity to consume insects (
Table 4). In
Table 4 the values of Infit and Outfit are equal to 0, in this case the model also considers extreme data (people who answered with the maximum or minimum to all items). These indices are not calculated since these people are excluded as the logarithm of the ratio
p/(1 −
p), at the basis of the model estimate, diverges [
73]. The average level of the measurements of non-extreme persons, i.e., those who did not respond to all items with the minimum or maximum value on the scale is −1.26. The reliability index in this case is 0.87. As shown in
Table 5, 69 people (17.2%) are extreme at the lowest levels, i.e., they responded with the lowest levels on the scale to all items. These people, representing 13.6% of the sample, were eliminated from the analysis because they had a bad adaptation index to the test (outfit or infit greater than 2.0).
Table 6 shows the model adaptation indexes for each item in more detail. It should be noted that all the point-measure correlations observed are close to those expected, plus the characteristic curves of the items ensure the optimal adaptation of the data and selected persons (item-persons) to the model. The items characteristic curves (in
Figure 1 the item characteristic curve—D11) allowed the suitability of the individual items to be tested to measure the propensity for entomophagy of each person. If the distance between the estimated curve and associated empirical curve is small, the capacity of the item to describe the phenomenon is greater.
The Rasch PCA of residuals looks for patterns in the part of the data due to randomness. Such a pattern is the “unexpected” part of the data that may be due, among other things, to the presence of multiple dimensions in the data [
71]. In the Rasch PCA of residuals, we are looking for groups of items sharing the same patterns of unexpectedness. In particular, the matrix of item correlations based on residuals is decomposed to identify possible “contrasts” (principal components) that may be affecting response patterns [
53]. Usually the contrast needs to have the strength (eigenvalue) of at least two items to be above the noise level: if the highest eigenvalue of PCA is around 2 or less the latent measure under investigation can be considered unidimensional. In this case, because the highest eigenvalue is 1.9462, we may conclude in favour of the uniqueness of the dimension investigated (
Figure 2).
In order to understand how to increase the propensity for insect consumption, it is useful to analyze
Table 7 where items are classified according to the degree of “difficulty”. The most “difficult” items are those that respondents have more difficulty in fully agreeing with. The items are: “Which insect(s) would you consider eating among the ones listed below” (D22.4—eat a beetle) with a maximum entomophagy propensity measurement of 2.77; “For which reasons would you decide to eat insects?” (D11.4), with a value of 2.27; “What feeling does this picture provoke in you? (Locusts with chocolate) (D31) with a value of 1.37. About item “How much would you be willing to pay for 10 g of ready-to-eat insects?” (D15M) for males this has a value of 1.17 which is higher than females (D15F) with value of 0.33. This means that for the same amount of insects (e.g., 1 hamburger) males have a higher propensity for entomophagy than females. This show that the propensity for entomophagy in males is much higher than that of females. This is in agreement with previous studies reporting that Italian females [
74,
75] and other European women in Belgium [
51,
76], The Netherlands [
77], Germany [
78], Hungary and Switzerland [
79] were less amenable to eating insect or insect-based food. This aspect will be further analyzed below. The easiest items on which to make a judgment of full agreement are respectively: “Are you interested in receiving more information about it?” (D20) with a measure of propensity for entomophagy at −3.60, answering “yes” to this question is the first step towards a growth of entomophagy. This is followed by “Would you be more willing to eat insects if you had a better understanding of the practice of entomophagy?” (D21) with a value of −2.70: answering “yes” to this question indicates a propensity for entomophagy higher than the item D20. According to previous study the growth of knowledge of a wide range of stimuli, including food, which entails a lower level of food neophobia [
80]. This shows that if a consumer is given enough information about a new food, such as insects, it increases the likelihood that they will consume it [
81,
82].
It is very interesting to analyze the item “Which insect(s) would you consider eating among the ones listed below? (D22: cricket 22.1; grasshopper 22.2; mealworm 22.3; beetle 22.4; silkworm 22.5; locust 22.6; ant 22.7; none 22.8). The willingness to consume one of the insects proposed by the item highlights the level of propensity for entomophagy of the respondent. Those who are inclined to consume a beetle (D22.4) showed a maximum propensity for entomophagy of 2.77, followed by the consumption of locusts (D22.6) with a score 0.88 lower than the previous one. This highlights the degree of exceptionality of item D22.4. The propensity scores for other insects are: 0.57 and 0.53 for mealworm (D22.3) and silkworm (D22.5) respectively. At a lower propensity score (−0.31) is the consumption of ants (D22.7), while the absolutely lowest propensity score (−1), which therefore indirectly indicates a greater consumption, is for crickets (D22.1) and grasshoppers around −1 (D22.2). The lower the value of the propensity score is, the greater is consumption.
The same considerations can be made for insect-based food products where there is a greater consumption for bakery products based on flour of (D24.1/2) with a propensity level of −1.55. The lowest propensity for insect food products is for protein powder, with a value of 0.40. The intake of insects in the form of bakery products greatly facilitates consumption because the insect is transformed into something “culturally” accepted while in the case of protein powder the cultural filter cannot be applied. cultural factors are known to influence food choices. Biological, psychological and socio-cultural factors are known to determine food choices. Culture, in particular, influences what is considered edible. Many people in Western countries rejecting the idea of entomophagy mainly for cultural reasons [
83].
The ranking of the items suggests the way that could be taken to lead an individual from the lowest level of the propensity for entomophagy scale (item D20) to the highest level (item D22.4: eating a beetle). Therefore, the steps to be taken are: (a) provide information to increase the propensity to consume; (b) stimulate the consumption of insect bakery products based on cricket or grasshopper flours; (c) include, in the human diet, insects that generally generate disgust, such as beetle. Despite the widely demonstrated benefits of using insects in human nutrition, the practice remains largely unacceptable in Western culture [
2,
84]. This lack of acceptance and consumption is attributed to the emotion of disgust associated with insect ingestion [
85]. Therefore, disgust drives the rejection of entomophagy. Several studies highlight that disgust emerges as the most common response, and represents the greatest barrier, to entomophagy [
75,
86,
87]. If a consumer does not have adequate information on the possibility that insects can be an integral part of his diet, he will be extremely wary and unlikely to consume them. It’s fundamental to proceed with an information campaign on the possible uses and positive effects of this new food behavior. Once the information is consolidated, the consumer will be invited to consume bakery products made with insect flours. In this way, there will be a positive association between the consumption of a traditional food, such as bread and cakes, produced using an innovative ingredient (insect flour). Once these positive experiences are consolidated, the last step can be taken as the direct consumption of insects.
Figure 3 allows a better understanding of the sequence of the various steps. On the right side are shown the items ordered according to the level of propensity for entomophagy. At the bottom are those showing a low propensity for entomophagy and at the top those showing a high propensity. At the bottom (and top), on a scale between −5 and +7 (values estimated by the program) the propensity for entomophagy is measured. The values (negative or positive) showed the level propensity for entomophagy. For example, the lowest level of propensity for entomophagy is for item D20 “Are you interested in receiving more information about it? and the answered is 0 (NO), and for item D21 “Would you be more willing to eat insects if you had a better understanding of the practice of entomophagy?” and the answered 0 (NO). The highest propensity level is for persons answering item D15 “How much would you be willing to pay for 10 g of ready-to-eat insects?” the equivalent of 2–3 hamburgers; to item D22 “Which insect(s) would you consider eating among the ones …?” “Beetle” is answered, and item D31 “What feeling does this picture provoke in you? (Locusts with chocolate)” answered “Appetite”.
Differential Item Functioning (DIF) analysis can be used to determine whether different subgroups respond differently to items, in particular whether gender deter-mines a greater or lesser difficulty items response. Easy items correspond to high propensity for entomophagy. In this survey the DIF analysis for males and females does not show anything relevant (test
t is within the limits −2.58, +2.58) in fact gender does not affect the perceived difficulty level of each item. However, there is an exception for the item “How much would you be willing to pay for 10 g of ready-to-eat insects? (D15) where the propensity to consume insects seems higher for males than females. Therefore, with the same response, males showed a higher propensity for entomophagy. For this reason, item D15 was treated as two separate items, one for males and the other for females (
Figure 4). As shown by several authors [
51,
88,
89] males, compared with females, were found to be more positive towards insects, less influenced by the species and level of processing of insects and more willing to eat insects for reasons of taste.
A further analysis dealt with the statistically significant correlations, in the sample, between the propensity for entomophagy, estimated with the model, with variables such sex, age, education and items that do not contribute to the estimation of the propensity to consume insects. The results of this analysis are shown in
Table 8 and following comments. This analysis confirmed the validity of the scale built by the items selected with the Rasch model, at least in the selected sample.
Again, specifying that the following results should be interpreted as descriptive measure of what observed in this convenience sample, and that the extension to the overall population would require further studies with random selected samples, the propensity level of females (coded as 2) is −2.61, below the sample average (−1.92) while that of males (coded as 1) is −0.89, above the average. Analysis of variance indicates that this difference is statistically significant with a p-value of 0.000. The analysis of variance between the different age groups and education of the sample does not show significant differences in propensity for entomophagy. The educational level is not a determinant variable in the propensity to consume insects as evidenced by Petersen et al. (2020). Instead, what is important is the positive experiences with insect food items that highlight the benefits of insect production and reduce the novelty of entomophagy are needed. The majority of respondents are Italian with an average propensity level of −2.20. The other nationalities are not present in a significant number to provide reliable information. In fact, the aim of this research is to validate the questionnaire and not describe the populations involved for which no representative samples are available. With regard to eating habits, it should be noted that vegetarians have a very low propensity for entomophagy (−3.13), while in omnivores the value is −1.95. The growth of vegan and vegetarian habits could therefore lead to a reduction in the propensity to accept insects as a food alternative.
Item D13 “Which of the foods below do you not eat (raw or cooked)?” concerns the consumption of various foods.
Table 9 shows the average propensity level to entomophagy of respondents who claim to consume and not consume such foods, with the p-value of the F-test of the difference between the two averages. This means that the consumption of certain foods is tested as a proxy for entomophagy propensity. It is noted that respondents who declared to consume one of these foods (with the exception of offal) showed a higher average propensity for entomophagy than those who declared not to eat them. The biggest difference in the average propensity for entomophagy between those who claim to eat a food and those who do not is found for clams. Regarding the consumption of offal there is no statistically significant difference in the propensity for entomophagy between respondents who declared they eat offal and those who do not. This particularity could be linked to the characteristics of offal considered by most consumers as an “extreme” food and consumed by few.
For item D16 “Which nutritional component do you think is the most present in insects?”, the differences are significant, compared to people who responded “protein” (first choice). This can be explained by people who claim to be vegetarian/vegan and are fundamentally opposed to any protein food that is not of vegetable origin. Answering item D27 “Would you eat animal products (meat, milk, etc.) produced by livestock fed with insects?” with “0—I do not consume animal products” has a very low average propensity to consume of −3.21. This confirmed that vegans and vegetarians tend to exclude insects from their diet. Therefore, the possible spread of vegan and vegetarian eating habits will be one of the main obstacles to the consumption of insects. The answer “1 = NO” has similarly a very low propensity equal to −3.45, while the answer “2 = YES” a propensity of −1.70. The differences are statistically significant. All the results of the one-way analysis of variance, that resulted significant, was related to 2 groups comparisons (sex and dichotomies “I eat/I do not eat” related to some kind of food), while analysis of variance for multiple groups variables, such as age and education, didn’t resulted significant and therefore there was not need of post-hoc tests or the so-called contrast analysis. In any case, as already underlined, the aim of this research was to validate the questionnaire and not describe the populations involved for which no representative samples are available.
4. Conclusions
In this research, the Rasch model was used, for the first time, to investigate the propensity to entomophagy by means of a questionnaire comprising 32 items designed to measure the respondents’ willingness to accept insects and/or insect-based products in their diet.
The value of the infit/outfit indices provided by the model ensure a good fit of the Rating Scale model. The indices, for both items and persons, have very high values that highlight the goodness of the scale used The research involved a large sample, among the largest in literature, although not representative of the Italian population. Despite this limitation, common to most papers dealing with entomophagy, the model provided useful information for a better understanding of the propensity for entomophagy.
However, the model provided other useful information for a better understanding the propensity for entomophagy. Indeed, the propensity is much higher in males than in females. This is in line with previous studies reporting that females were less likely to eat insect or insect-based food.
The analysis confirmed that entomophagy has mainly cultural bases and does not correlate with the level of education [
90]. This result suggests that the first step in increasing consumer confidence is to disseminate the positive effects of eating insect-based foods. Several studies showed that the barrier to consume food made with insects is essentially represented by Western culture, which does not consider insects as a possible form of food [
32,
91]. However, overcoming attitudinal barriers, such as negative taste expectations and neophobic reactions, through familiarity enhancement and taste education is paramount to establishing insects as a food source [
78]. The process leading the consumer to accept this new source of food through an insect product with the addition of traditional products (meat, chocolate, etc.), can increase people’s willingness to eat insects. This is the first step needed to make this new source of food a part of consumer’s food culture.
This process, present in Western societies, could be eliminated by creating greater familiarity with insect-based food and demonstrating the nutritional and health benefits that could break down the barrier to consumption and market development. The future will therefore depend on consumer reactions and socio-cultural characteristics, perception of risk to health and production technologies [
82].
The Food and Agriculture Organization of United Nations supports the production of edible insects as a promising more sustainable source of nutrients for the increasing demand for animal-derived products by the growing world population than traditional animal products. Even if insects are part of the diet of more than two billion people worldwide, the practice of eating insects (entomophagy) raises challenging questions for the Western countries where this practice is not a habit.
Finally, we argue that the Rasch model allowed us to better understand the steps that should be taken to improve the consumption of insects-based foods in Western countries.