The median length of the experiment was 12.4 minutes. Consistent with census data, 53.1% of the 4385 participants were female. Mean age was 44.91, SD=17.86; 35.6% reported children under the age of 18 and 21.6% reported children under age 5 lived in their households. When asked “who does the grocery shopping for the household,” 54.3% reported doing “all of it,” 18.2% “most of it,” 17.4% “about half of it,” 6.9% “some of it,” 2.5% “someone else does all of it,” and 0.7% preferred not to answer. Additional sociodemographic characteristics of the participants are shown in Table 1.
Of the 1452 participants randomly assigned to view a beef product, 96.2% reported having eaten one or more meals containing beef in the 12 months prior to the survey. Of the 1469 participants who viewed a chicken product, 97.5% had eaten at least one meal containing chicken in the prior 12 months. Of the 1464 participants assigned to see a salmon product, 77.0% indicated that they had eaten at least one meal containing salmon in the prior year.
Asked to rate their familiarity with the product they saw, the participants were moderately to very familiar with Chicken Breasts (M=3.88, SD =1.15), beef burgers (M=3.53, SD =1.31), Salmon Fillets (M=3.20, SD =1.30), and Beef Filets (M=3.02, SD =1.30). They only were slightly to moderately familiar with Chicken Burgers (M=2.68, SD =1.32) and salmon burgers (M=2.33, SD =1.34) [Scale: 1 not at all; 2 slightly; 3 moderately; 4 very; 5 extremely].
3.1 Criterion A – Ability to distinguish from conventional products
An essential function of a new common or usual name is to signal that the product bearing it is different from the products with which consumers may already be familiar. Z-tests for equality of proportions using a Bonferroni correction were used to examine differences in the ability of each of the common or usual name to communicate to the participants that the product was different from its conventional counterpart (Tables 2a-4b).
Overall, there were different patterns of results depending on the protein (beef, chicken, salmon), and the form of the product (whole cut or burger) shown to the participants. As shown in Table 2a, for Beef Filets, the common name “Cultivated” performed most poorly in signaling that the product was “Neither Grass-Fed nor Grain-Fed Beef.” Only 57.0% of the participants correctly identified the product as being different from conventional beef products and one-third (33.9%) mistakenly believed that the product was “Grass-Fed Beef.” In contrast, 76.6% of those who saw the Control product (with no term tested) reported that the beef fillets were “Neither Grass-Fed nor Grain-Fed Beef” and only 13.7% thought that the product was “Grass-Fed Beef.” The other terms tested performed at least as well as the Control in signaling that the product was “Neither Grass-Fed nor Grain-Fed Beef.”
For Beef Burgers, the pattern of results was different. There were no statistically significant differences among any of the common or usual names tested or the Control with respect to the proportions of participants who thought that the products were “Neither Grass-Fed nor Grain Fed” (Table 2b).
There were also no differences among the names tested in signaling that the Chicken Breasts were “Neither Free-Range nor Raised Indoors” and none outperformed the Control (72.5%) (Table 3a). Yet, when testing the terms on packages of Chicken Burgers, neither “Cultured” (58.6%) nor “Cultivated” (58.5%) performed as well as the Control (76.2%) in signaling that the Chicken Burgers were “Neither Free-Range nor Raised Indoors” (Table 3b).
As expected, for Salmon Fillets (Table 4a), the terms “Cultivated” and “Cultured” performed least well. Only one-third (33.1%) of those who saw the term “Cultivated” thought it was “Neither Wild-Caught nor Farm-Raised” and 41.7% thought it was “Farm-Raised.” Fewer than half (49.2%) of the participants who saw the term “Cultured” responded that the product was “Neither Wild-Caught nor Farm-Raised,” while 30.6% thought they were “Farm-Raised.” In contrast, 70.8% of those who saw the term “Cell-based” correctly identified the products as “Neither Wild-Caught nor Farm-Raised,” as did 56.6% of those who saw “Cell-cultured” and 58.0% of those who saw the term “Cell-Cultivated.”
The terms “Cultivated” and “Cultured” also performed poorly for Salmon Burgers (Table 4b). “Cultivated” indicated to less than one-third of the participants (32.5%) that the salmon was “Neither Wild-Caught nor Farm-Raised” and nearly half (48.0%) thought the salmon was “Farm-Raised.” “Cultured” signaled to only 46.1% of the participants that the salmon was “Neither Wild-Caught nor Farm-Raised” and 27.8% thought it was “Farm-Raised.” In contrast, 66.4% of those who saw the term “Cell-Based,” 66.1% of those who saw “Cell-Cultured,” and 60.9% of those who saw “Cell-Cultivated” correctly identified the products as “Neither Wild-Caught nor Farm-Raised.”
A two-way ANOVA examining the effects of the names tested and the products tested on the confidence the participants had in their answers regarding whether the product was a conventional product found no significant interaction effect (F(5, 4349) = 1.238, p =.192). However, there was a main effect of name tested (F(5, 4349) = 4.816, p < .001, ηp2 = .006). The participants who saw the Control (M=2.97, SD =1.47) and the products labeled as “Cultured” (M=2.98, SD =1.37) and “Cultivated” (M=3.01, SD =1.32) were least confident in their answers [Scale: 1 not at all; 2 slightly; 3 moderately; 4 very; 5 extremely confident]. Participants were significantly more confident in their answers when the products they saw were labeled as “Cell-Cultivated” (M=3.22, SD =1.40) and “Cell-Cultured” (M=3.21, SD =1.38) than when they were labeled as “Cultured” or “Cultivated.” The confidence of those who saw products labeled as “Cell-Based” (M=3.07, SD =1.36) was not significantly different from the confidence of those who saw any of the other terms.
There was also a main effect of the product tested (F(5, 4349) = 4.445, p < .001, ηp2 = .005). Those who saw the Chicken Burgers (M=2.91, SD =1.40) were less confident in their answers than those who saw the Salmon Burgers (M=3.23, SD =1.33) or the Salmon Fillets (M=3.14, SD =1.34). Confidence was not significantly different among those who viewed the Beef Burgers (M=3.09, SD =1.38), Beef Filets (M=3.07, SD =1.44) or the Chicken Breasts (M=3.02, SD =1.42), Salmon Burgers, or Salmon Fillets.
One-way ANOVAs indicated no main effect of name tested on how likely the participant indicated they would be to search for more information about the product on the Internet (F(5, 4349) = 1.42, p = .213) or to scan a QR code for more information about the product (F(5, 4349) = .997, p = .430).
3.2 Criterion B – Signal the presence of potential allergens
The ability to signal potential allergenicity is also a critical regulatory criterion. The participants were asked, “If you are allergic to Beef/Chicken/Salmon, how safe is it for you to eat these Beef Filets/Beef Burgers/Chicken Breasts/Chicken Burgers/Atlantic Salmon Fillets/Salmon Burgers?” [Scale: 1 very unsafe; 2 moderately unsafe; 3 somewhat unsafe; 4 neither safe nor unsafe; 5 somewhat safe; 6 moderately safe; 7 very safe]. A two-way ANOVA showed no interaction effects between name tested and protein tested (beef/chicken/salmon) (F(10, 4367) = .215, p = .995). There was a main effect of the protein tested (F(2, 4367) = 11.71, p < .001, ηp2 = .005). The salmon products were judged less safe to consume by those allergic to salmon (M=2.82, SD =2.11) than consumption of the beef products by those allergic to beef (M=3.13, SD =2.15) or consumption of the chicken products by those allergic to chicken (M=3.16, SD =2.14).
There was also a main effect of the name tested (F(5, 4367) = 7.548, p < .001, ηp2 = .009). All the names and the Control appropriately signaled that it was moderately to somewhat unsafe to eat the products if one were allergic to the protein from which they were made (Table 5). However, the Control (M=2.73, SD =2.13) and the products with the terms “Cultivated” (M=2.87, SD =2.14) were seen as least safe to consume by those allergic to the protein.
3.3 Criteria C and D - Not be viewed as disparaging of cell-based or conventional products
The participants were asked to carefully examine the package of seafood shown to them and asked to type their response to the question, “What is the first thought, image, or feeling that comes to mind when seeing this package?” They were then asked to look at the package a second time and to record the second thought, image, or feeling that came to mind. Each of the responses was coded using one of the categories developed by Hallman and Hallman (2020) (see Tables S1 and S2 in the supplemental materials). Each response was independently coded by two trained researchers, with any discrepancies resolved by consensus. After recording their open-ended responses, each participant rated how positive or negative their thought, image, or feeling was, using a scale ranging from 1 extremely negative to 7 extremely positive. The participants were asked to look at the package a third time and to record how positive or negative their overall reactions were. A MANOVA examining the effects of name tested and product tested on all three ratings as dependent measures found main effects of name, F(15, 11998) = 2.662, p < .001; Wilk's Λ = 0.991, ηp2= .003, and product, F(15, 11998) = 4.016, p < .001; Wilk's Λ = 0.986, ηp2= .005, but no interaction F(75, 12993) = 1.150, p = .117; Wilk's Λ = 0.980, ηp2= .007.
As shown in Table 6, the first and second thoughts, images, and feelings and overall reactions associated with the control products, and those products labeled with the terms “Cultivated” and “Cultured,” were as positive as those labeled with the terms “Cell-Based,” “Cell-Cultured.” “Cell-Cultivated” was viewed least positively. With respect to the products, the Beef Filets garnered the most positive responses, while the Chicken Burgers received the least positive reactions (Table 7).
A MANOVA was used to explore the effects of name and product tested on the dependent measures described below. The analysis showed main effects of name, F(55, 2083) = 6.640, p < .001; Wilk’s Λ = 0.920, ηp2= .017, and product, F(55, 2083 = 12.696, p < .001; Wilk’s Λ = 0.854, ηp2= .031, but no interaction effect, F(275, 47828) = 1.108, p = .105; Wilk’s Λ = 0.932, ηp2= .006. We therefore focus on the main effects of name tested.
Interest in Tasting, Likelihood of Purchasing, Ordering, Serving to Guests
There was a significant effect of name on interest in tasting the products (F(5,4348) = 4.426, p < .001, ηp2 = .005). The participants were slightly to moderately interested in tasting all the products. They were equally interested in tasting the Control and the products labeled as “Cultured,” “Cultivated” and “Cell-Cultured,” and less interested in tasting the products labeled as “Cell-Based” and “Cell-Cultivated” (Table 8).
There was a significant effect of name on reported likelihood to purchase the products in the next six months (F(5,4348) = 4.117, p < .001, ηp2 = .005). Participants were “neither likely nor unlikely” to “slightly likely” to purchase all of the products. They were most likely to purchase the Control products, those labeled as “Cultured,” “Cultivated,” and “Cell-Cultured,” and slightly less likely to purchase those labeled as “Cell-Based,” or “Cell-Cultivated.” (Table 9).
There was a significant effect of name tested on likelihood to order the products in a restaurant (F(5,4348) = 4.212, p < .001, ηp2 = .005). Participants were “neither likely nor unlikely” to “slightly likely” to order the products. They were most likely to order the “Cultured” and Control products and those labeled as “Cultivated” and “Cell-Cultured” slightly less likely to order those labeled as “Cell-Based” and “Cell-Cultivated” (Table 10).
There was a significant effect of name on likelihood to serve the products to guests in the next six months (F(5,4348) = 5.052, p < .001, ηp2 = .006). Participants were “neither likely nor unlikely” to “slightly likely” to serve the products to guests. They were most likely to serve the Control products and least likely to serve those labeled as “Cell-Based” (Table 11).
Safe to Eat
All the names appropriately signaled that if one is not allergic to the protein from which they are made, it is somewhat to moderately safe to eat the products “Cultivated” (M=5.98, SD =1.40), “Cultured” (M=5.95, SD =1.52), Control (M=5.94, SD =1.49), “Cell-Cultivated” (M=5.83, SD =1.45), “Cell-Based” (M=5.79, SD =1.54), “Cell-Cultured” (M=5.75, SD =1.55) [Scale: 1 very unsafe to 7 very safe]. However, the products with the term “Cultivated” were seen as safer than those labeled as “Cell-Cultured” (F(5,4348) = 3.042, p = .010, ηp2 = .003).
Natural, Organic, Genetically Modified
There was a main effect of name tested on participant perceptions of the product’s naturalness [Scale: 1 very unnatural to 7 very natural] (F(5,4348)) = 20.887 p < .001, ηp2 = .023). The Control (M=5.25, SD =1.51) was perceived as the most natural product. The products labeled as “Cultivated” (M=5.00, SD =1.63), and “Cultured” (M=4.93, SD =1.65) were seen as equally natural, and both were viewed as more natural than products with the terms “Cell-Cultivated” (M=4.61, SD =1.84), “Cell-Based” (M=4.60, SD =1.81) and “Cell-Cultured” (M=4.51, SD =1.85), which were seen as equally natural.
There was no main effect of name tested on participant perceptions of the likelihood that the product is organic [Scale 1 extremely unlikely to 7 extremely likely] (M=4.31, SD =1.83), (F(5,4348)) = 0.735 p = .597). There was a main effect of name tested on participant perceptions of the likelihood that the product is genetically modified [Scale 1 extremely unlikely to 7 extremely likely] (F(5,4348)) = 34.827 p < .001, ηp2 = .039). Products with the terms “Cell-Cultivated” (M=5.30, SD =1.64), “Cell-Based” (M=5.24, SD =1.60) and “Cell-Cultured” (M=5.23, SD =1.58) were seen as more likely to be genetically modified than products with the terms “Cultured” (M=4.81, SD =1.60) and “Cultivated” (M=4.79, SD =1.63). The Control (M=4.41, SD =1.66) was seen as the least likely to be genetically modified.
Nutritious, Healthy
After being shown the enlarged nutrition facts label, the participants were asked how nutritious they thought the products are [Scale: 1 not at all; 2 slightly; 3 moderately; 4 very; 5 extremely] and how healthy they are [Scale: 1 extremely unhealthy – 7 extremely healthy]. There was no main effect of name tested on perceptions of nutritiousness (M=3.31, SD =1.14) (F(5,4348)) = 0.318 p =.902) or on perceptions of healthiness (M=4.98, SD =1.60) (F(5,4348)) = 0.436 p =.824). Overall, the products were seen as moderately nutritious and neither healthy nor unhealthy. Perceptions of nutritiousness and healthiness were strongly correlated r(4383) = .74, p < .001.
Taste
There was a significant main effect of name tested on how the participants think the product tastes [Scale: 1 extremely bad – 7 extremely good] (F(5,4348)) = 3.254 p = .006, ηp2 = .004). Each product was thought to taste slightly to moderately good. Those labeled as “Cultured” were thought to taste slightly better (M=5.33, SD =1.58) than those labeled as “Cell-Based” (M=5.07, SD =1.62). Post hoc tests detected no other differences; Control (M=5.31, SD =1.65), “Cultivated” (M=5.29, SD =1.56), “Cell-Cultured” (M=5.20, SD =1.61), “Cell-Cultivated” (M=5.12, SD =1.60).
Likelihood to Recommend to Pregnant Women and to Children
A MANOVA found main effects of product tested F(10, 8696) = 9.985, p <.001; Wilk's Λ = 0.977, ηp2= .004 and name tested F(10, 8696) = 3.084, p <.001; Wilk's Λ = 0.993, ηp2= .011, on the likelihood that the participant would recommend that pregnant women eat the product and that children eat the product. There was no interaction F(50, 8696) = 1.28, p = .026; Wilk's Λ = 0.984, ηp2= .008.
Examining the main effect of name on likelihood to recommend that those who are pregnant eat the products, (F(5,4348)) = 4.043 p = .001, ηp2 = .005), the participants reported that they would be “neither likely nor unlikely” to recommend the products be eaten by pregnant women (Table 11). They were most likely to recommend that the Control products be consumed by those who are pregnant and less likely to recommend the “Cell-Cultured”, “Cell-Based” and “Cell-Cultivated” products.
Similarly, the participants reported that they would be “neither likely nor unlikely” to recommend the products be eaten by children (Table 12) (F(5,4348)) = 4.578 p < .001, ηp2 = .005). They were most likely to recommend that the Control products be consumed by children and less likely to recommend the “Cell-Cultured” and “Cell-Based” products.
3.4 Criterion E – Be seen as an appropriate term
After reading the explanation of the meaning of the term they had been randomly assigned to see on the product packages, the participants were asked how familiar they were with the idea of producing just the parts of beef/chicken/salmon that people eat, instead of raising (or catching) them whole and harvesting them. Of the 3,652 participants not in a Control condition, 54.4% reported that they were “not familiar at all,” 12.9% “slightly familiar,” 13.9% “moderately familiar,” 9.6% “very familiar,” and 9.1% “extremely familiar” with the idea of producing beef/chicken/salmon products in this way.
They were asked how appropriate the term was for describing this new way of producing just the parts of beef/chicken/salmon that people eat. All the names were judged to be “neither appropriate nor inappropriate” to “slightly appropriate.” There was a main effect of name in judgements of appropriateness of the term (F(4,3647)) = 3.802 p = .004, ηp2 = .004). As shown in Table 13, the term “Cultivated” was seen as the least appropriate term and as significantly less appropriate than both “Cell-Cultivated” and “Cell-Based”.
The participants who saw a beef product were asked how clear the term was in communicating that the beef was not “Grass-Fed,” and that it was not “Grain-Fed.” Those who saw a chicken product were asked how clear the term was in communicating that the chicken was not “Free-Range,” and that it was not “Raised Indoors.” Those who saw a salmon product were asked how clear the term was in communicating that the chicken was not “Wild-Caught,” and that it was not Farm-Raised.
There was a main effect of name tested on how clear the term was in communicating that the beef products were not “Grass-Fed” (F(4,1208)) = 3.875 p = .004, ηp2 = .013) and a main effect of name on clarity that the products were not “Grain-Fed” (F(4,1208)) = 3.788 p = .005, ηp2 = .012). Products with the term “Cell-Cultured” were seen as clearer in communicating that the products were not “Grain-Fed” than products labeled with either “Cultivated” or “Cultured” (Table 14).
There was no main effect of name tested on how clear the term was in communicating that the chicken products were not “Free-Range” (F(4,1222)) = 1.274 p = .278). All names were judged to be “neither clear nor unclear” to “slightly clear” (M=4.53, SD =2.06) [Scale: 1 Extremely Unclear to 7 Extremely Clear]. There was similarly no main effect of the name in terms of clarity in communicating that the products were not “Raised Indoors” (F(4,1222)) = 1.274 p = .224) (M=4.43, SD =2.08).
Each name was seen as equally clear in communicating that the salmon products were not “Wild-Caught” (M=4.60, SD =2.04) (F(4,1212)) = 2.398 p = .048). There was a main effect of name on clarity in communicating that the products were not “Farm-Raised” (F(4,1212)) = 6.128 p < .001, ηp2 = .020). The term “Cultured” was seen as least clear in conveying that the products were not from aquaculture, while the terms containing the word “cell” were seen as clearer in communicating that the products were not “Farm-Raised” (Table 15).
The participants were asked how clear the name was in communicating that the product they had seen was not made from plants. There was a main effect of name tested (F(4,3647)) = 4.322 p = .002, ηp2 = .005). As shown in Table 16, all names tested were seen as “neither clear nor unclear” to “slightly clear”. The term “Cell-Cultured” was seen as clearer in communicating that the products were not plant-based than the terms “Cell-Based,” “Cultivated” and “Cultured.”
Participants viewing the novel beef/chicken/salmon products were asked how much they agreed or disagreed that they should be sold in the same section of the supermarket as those that are “Grass-Fed and Grain-Fed,” “Free-Range and Raised Indoors,” or “Wild-Caught and Farm-Raised” [Scale: 1 Strongly disagree to 7 Strongly Agree]. Univariate ANOVAs showed no main effects of name tested for beef (F(4,1203)) = 1.626 p = .760), chicken (F(4,1212)) = 1.87 p = .114), or salmon products (F(4,1212)) = 1.506 p = .198). The participants “Neither Agree nor Disagree” to “Somewhat Agree” that the beef (M=4.51, SD =1.86), chicken (M=4.32, SD =1.89), and salmon products (M=4.49, SD =1.85) should be sold in the same section of the supermarket as their conventional counterparts.
3.5 Consumer perceptions after learning the meaning of the term.
After learning the meaning of the term they had seen, the participants were shown the same package a final time, and asked for their overall reactions, their interest in tasting the product, and the likelihood they would purchase the product in the next six months if it were available in their grocery store. A MANOVA examining these dependent measures found that there were no main effects of name tested after explaining its meaning, F(32, 13329) = 1.136, p =.274; Wilk's Λ = 0.990, ηp2= .003. This suggests that any potential marketing advantages a name may initially have are likely to disappear after consumers achieve greater awareness and understanding of the products and the technology used to produce them.
To illustrate this, a repeated measures ANOVA showed a main effect of explaining the term on overall reactions (F(1,3647)) = 400.752 p <.001, ηp2 = .099), with significant declines in how positive the participant’s reactions were to the products after reading the explanation (Figure 2). There was also an interaction effect of the explanation and the name the participant saw (F(4,3647)) = 15.15 p <.001, ηp2 = .014). Prior to the explanation, the overall ratings for products with the term “Cultivated” were significantly higher than those for products with the other names (Table 6). After the explanation, overall ratings for products labeled with the term “Cultivated” dropped to the lowest measured among the names (though not significantly different from them). The same pattern was observed with respect to the main effect of the explanation on interest in tasting the products (F(1,3647)) = 308.323 p <.001, ηp2 = .078), and the interaction effect of the explanation and the name the participant viewed (F(4,3647)) = 7.065 p <.001, ηp2 = .008) (Figure 3). This pattern was repeated in the main effect of the explanation on likelihood of purchasing the product in six months if available in the participant’s grocery store (F(1,3647)) = 416.206 p <.001, ηp2 = .102), and the interaction effect of the explanation and the name the participant viewed (F(4,3647)) = 10.032 p <.001, ηp2 = .011) (Figure 4). Interest in tasting and likelihood to purchase products labeled as “Cultivated” dropped significantly more after reading the explanation than was the case for the other terms.
3.6 Discussion
As expected, consumer awareness of the concept of producing just the parts of beef/chicken/salmon that people eat, instead of raising (or catching) them whole and harvesting them, remains very low. Most of the participants (67.4%) reported that they were “not familiar at all” or only “slightly familiar” with the idea. Therefore, the common or usual name chosen must convey significant new information to uninformed consumers. This includes communicating “the basic nature of the food or its characterizing properties or ingredients” (21CFR102.5), and what distinguishes it from other foods. The name chosen must also do so on its own; without the benefit of additional explanatory labeling text or other supporting materials.
None of the names tested in this experiment are part of existing labeling schemes for beef, chicken, or salmon products. Although the methods of production (“Wild-Caught” and “Farm-Raised”) are required to appear on seafood labels (7CFR60.300), no similar regulations for labeling the methods of livestock production exist for meat and poultry products. Labels such as “Grass-Fed” and “Free-Range” are voluntary and the products bearing them are often sold at a price premium, with each making up about one percent of the market for beef or chicken. Therefore, the signal sent by the proposed names that this new method of production is different from conventional beef, chicken, and salmon is likely contingent on a consumer’s familiarity with the existing labels for those conventional products.
It is instructive to examine the Control conditions for each product. Because conventional salmon products are already labeled as “Wild-Caught” or “Farm-Raised,” the inclusion of no common or usual name (the Control condition) leads many consumers to assume that the products are “Wild-Caught.” This may be their default assumption if products are not specifically labeled as “Farm-Raised.” The data from this experiment shows that introduction of the term “Cultivated” and the absence of the “Wild-Caught” label signals that the products are “Farm-Raised.”
Similarly, in the Control condition for Beef Filets and Beef Burgers, the default assumption appears to be that in the absence of a label that specifically indicates that the products are “Grass-Fed” or “Grain-Fed,” the products are neither. When “Cultivated” is on the label of the Beef Filets, it is mistaken by one-third of consumers as meaning that the product is “Grass-Fed.” Yet, when selecting Beef Burgers, consumer choice between “Grass-Fed” and “Grain-Fed” beef may not be as salient, so any additional label that consumers have not seen before may signal that the product is different from those they usually purchase. This would explain why all of the names performed as well as the Control in signifying that the Beef Burgers were “Neither Grass-Fed nor Grain-Fed.
In the Control condition for Chicken Breasts and Chicken Burgers, the absence of a label that specifically indicates that the products are “Free-Range” or are “Raised Indoors,” might cue a default assumption that the product is neither. Yet, “Cultured” and “Cultivated” performed more poorly than the control in signaling that the Chicken Burgers were “Neither Free-Range nor Raised Indoors.” Though not measured in this study, it is likely that widespread lack of familiarity with conventional (indoor) chicken production systems by consumers33-34 and their lack of exposure to labeled “Free-Range” alternatives in supermarkets made those categories less salient to the participants.
Given the significant success of plant-based meats in the US35, the other salient category for consumers across beef, chicken, and salmon products may be “Plant-Based.” Across the products analyzed, “Cell-Cultured” and “Cell-Cultivated” were viewed as clearest in conveying that the products were not plant-based.
The results of this study are consistent with Hallman & Hallman12 in showing that the common or usual names “Cultured” and “Cultivated” inadequately communicated to consumers that the Salmon Fillets and Salmon Burgers presented were different from “Wild-Caught” and “Farm-Raised” Salmon, signaling instead to many consumers that the products were “Farm-Raised.” They therefore fail to meet Criterion A, suggesting that adoption of the terms “Cultured” and “Cultivated” would be problematic with respect to the implementation of a single name across meat, poultry, and seafood products. The results are also consistent with Hallman & Hallman12 in showing that terms containing the word “cell” (“Cell-Based,” “Cell-Cultured,” “Cell-Cultivated”) were more effective in conveying to consumers that the products with those labels are different from conventional seafood products.
The participants were least confident in their answers regarding whether the beef, chicken or salmon product was different from conventional products when they saw them labeled as “Cultured” or “Cultivated.” The terms “Cultivated” and “Cultured” were also seen as least clear in communicating that the beef and salmon products were different from their conventional counterparts. Consistent with this, the term “Cultivated” inadequately conveyed that the Beef Filets were different from conventional products, signaling to more than one third (33.9%) that the products were “Grass-Fed.” These results, particularly in comparison to the performance of the Control products (with no common or usual name) suggest that "Cultivated" may be misleading to consumers, failing to meet Criterion A for both beef and salmon products. Though not tested in conjunction with shellfish products in this experiment, the term “Cultured” is also commonly used to describe the production of farmed shellfish, making it an inappropriate term to identify shellfish that have not been farmed. “Cultured” is also used to identify fermented dairy products, which may also be problematic to the adoption of a single term across all protein products.
Each name appropriately signaled to consumers that if they are allergic to beef/chicken/salmon it is “somewhat” to “moderately unsafe” to eat these novel products, meeting Criterion B. While the effect size was small, it should be noted that in comparison to the Control, the addition of a common or usual name to the label signaled that the product was slightly safer for those allergic to eat. This suggests that allergen warnings may need to be highlighted on packages of products made directly from the cells of animals.
None of the names were perceived as inappropriate for communicating the idea of producing just the parts of animals that eat instead of raising (or catching) them whole and harvesting them, meeting Criterion E. While the difference was small, the term “Cultivated” was seen as least appropriate.
By design, none of the names tested were disparaging of other products (Criterion C), and the coding of the open-ended responses after seeing the products did not reveal a clear pattern of negative thoughts, images, or feelings associated with any of the names tested. Before learning the meaning of the names, the participants were slightly to moderately interested in tasting all the products, but least interested in tasting the products labeled as “Cell-Based” and “Cell-Cultivated.” The participants reported that they were “neither likely nor unlikely” to “slightly likely” to purchase the products. They reported that they were least likely to purchase those labeled as “Cell-Cultivated,” and least likely to serve “Cell-Based” products to guests.
None of the names tested suggested that the products were unsafe to eat or unnatural, although those labeled as “Cell-Cultivated,” “Cell-Based,” and “Cell-Cultured” were seen as less natural than those labeled as “Cultivated” or “Cultured.” No organic standards exist for these products, and none of the names implied that the products were likely to be organic.
However, products labeled with “Cell-Cultivated,” “Cell-Based,” and “Cell-Cultured” were seen as more likely to be genetically modified. Some production methods for these products may rely on inputs that are genetically modified36. However, those products that do not involve genetic modification may wish to provide additional labeling indicating this, as may be permitted by regulation.
None of the names tested influenced perceptions of how nutritious or how healthy the products were perceived. All were perceived as moderately nutritious and neither healthy nor unhealthy. Each product was imagined to taste slightly to moderately good, though those labeled as “Cultured” were thought to taste slightly better than those labeled as “Cell-Based.” The participants were more likely to recommend that people who are pregnant eat the Control products than the “Cell-Cultured,” “Cell-Based” and “Cell-Cultivated” products. They were also more likely to recommend that children eat the Control products than those labeled as “Cell-Cultured” and “Cell-Based.” They were also not opposed to selling the novel beef, chicken, and salmon products in the same section of the supermarket as their conventional counterparts. Therefore, each name meets Criterion D.
The results also show that after reading an explanation of the meaning of the terms, the main effects of the common or usual name on key consumer acceptance measures disappear, including overall reactions to the product, interest in tasting, and likelihood to purchase the products if available in their grocery store. This suggests that once consumers have greater familiarity and understanding of these novel products and the processes used to create them, any of the initial marketing advantages of the names “cultivated” and “cultured” (some of which may be based on consumer misperceptions that they are conventional products) are likely to vanish. Moreover, the differences between initial overall reactions to products labeled with these terms and those after the explanation of the term may indicate the possibility of a consumer “backlash” related to learning that their initial perceptions of the nature of the product was incorrect. Not directly measured in this experiment, this effect may be worthy of future research.
In choosing the best name among “Cell-Based,” “Cell-Cultured” and “Cell-Cultivated,” it should be noted that the differences in the means of many of the key dependent measures and their associated effect sizes are quite small. There were no statistically significant differences among the three in their ability to signal that the products were different from conventional products or in signaling allergenicity. There were also no statistically significant differences among the three with respect to key consumer acceptance measures, including initial overall reactions, interest in tasting, likelihood to purchase, likelihood to purchase in a restaurant and likelihood to serve to guests, likelihood to recommend that pregnant women and children eat the products, the perceived appropriateness of the term, perceived clarity in communicating that the product is different from conventional products, or made from plants. The best name of the three is likely to be the one that achieves consensus across the meat, poultry, and seafood sectors and is actively promoted by all.
However, examining its ranked position with respect to both the Control conditions and those of “Cell-Based” and “Cell-Cultivated,” the overall pattern of results across the beef, chicken, and salmon products suggests that the term “Cell-Cultured” may be most advantageous to adopt. It performs well in communicating that the products are different from conventional products and in communicating allergenicity, thereby meeting the two key regulatory criteria. With respect to consumer acceptance, the participants indicated that they are as interested in tasting “Cell-Cultured” products, as likely to purchase them, as likely to order them in a restaurant, and as likely to serve them to guests as the Control products. Given that meat and poultry products in the US are not required to have labels declaring their production methods, the Control packages shown in this study represent packages of conventional meat without any voluntary labeling with respect to production method. It is against those conventional meat and poultry products which the “Cell-Cultured” products would compete.
As with any experiment, this study has limitations. The participants saw high-resolution images of the products and were not able to physically interact with them as they might in a grocery store. They packages were also seen in isolation, without the context of having other products on the same grocery shelf to which they might be compared. Tests of the packages and common or usual names under realistic shopping conditions would add to the strength of this study.