Evolutionary thinking among biology students in a third world country

The evolution–creationism controversy has been a longstanding issue ever since the publication of Darwin’s seminal work, On the Origin of Species, in 1859 (Moore et al. 2009). The opposed perspectives of creationism and evolution are associated with incompatible views on culture and religion as well as science (Miller et al. 2006; Moore et al. 2009; Scott 2009; Heddy and Nadelson 2012).

Despite being contentious subjects, creationism and evolution are not dichotomic in people’s beliefs. As creationism displays several facets, there is a continuum between these two extremes, depending on the degree of science acceptance (Scott 2009). Among several viewpoints, for instance, classic creationism connotes that the creation event occurred relatively recently, approximately 10,000 years ago, and life has been created as it is today. Alternatively, various intermediate ideologies consider Earth to be ancient and the Creator to be responsible for the start and for changes along the process (sensu Scott 2009). However, scientific support is lacking for these and all other forms of divinely guided processes. On the other hand, naturalist evolutionary theory is one of the most robust and important in history, without being credibly falsified so far (Gould 2002; Mayr 2007; Scott 2009). It provides a materialist interpretation of issues such as the diversification, interrelatedness, and distribution of life on Earth. Evolutionary theory has been continuously supported by science, including natural selection, adaptation, selfish selection, mutations, genetic variation, gene flow, genetic drift, competition, fitness, extinction, and fossil record, offering a high explanatory and predictive power of natural processes (Gould 2002; Mayr 2007; Scott 2009).

Opinion surveys have been a valuable tool for investigating evolutionary thinking across different publics and localities. For instance, acceptance of evolution may be negatively related to social dysfunction indexes (Paul 2009) and income inequality (Rees 2009), and positively related to logical thinking skills (Lawson and Weser 1990; Lawson and Worsnop 1992). However, evolutionary thinking is most commonly positively related to education and scientific literacy (Miller et al. 2006; Rice et al. 2011; Sherkat 2011; Wiles and Alters 2011; Heddy and Nadelson 2012, 2013; Lloyd-Strovas and Bernal 2012) and negatively to religiosity (Lawson and Worsnop 1992; Sinclair et al. 1997; Larson and Witham 1998; Miller et al. 2006; Sherkat 2011; Heddy and Nadelson 2012, 2013; Clément 2015), suggesting that religiosity drives resistance to evolution (Coyne 2012).

However, evolution is expected to be predominant in some groups such as college biology students, due to their close contact with this theory and high level of education (Dias et al. 2012). Nevertheless, there is a lack of comparative studies with statistical analysis, sufficient sample sizes, and large geographic range, which hinders the understanding and decision-making of how and where to intervene (Nehm 2006; Lloyd-Strovas and Bernal 2012). In this sense, despite the assumptions of evolutionary predominance, more information is necessary, especially in less studied groups, such as postgraduate students. Finally, most researches concerning public opinions in evolution are concentrated in developed, rich countries (mainly in Europe and the USA), with few studies in the underdeveloped and emerging ones.

Brazil boasts the major research and graduate system in Latin America and the Caribbean (Andrade and Galembeck 2009). It is currently the 24th in global research, with areas that include biology such as “Earth & Environmental Sciences” and “Life Sciences” together comprising approximately ¼ of the top research in the country (Nature Index 2016). Thus, it is important to understand students’ perceptions on evolution, as they will be the future researchers and teachers, playing an important role in public opinion formation. Taking Brazil as a model system, we conducted a systematic survey on biology students’ evolutionary opinions, contrasting creationism, divinely guided evolution and naturalist evolution (without divine intervention). Although Brazil has a large Catholic and, more recently, Evangelical, Pentecostalist and Adventist influences (IBGE 2012), we hypothesized that (1) biology students are more inclined to naturalist evolution. Despite Brazil’s large territorial extension, we expected that (2) there are no regional effects on students’ opinions once all of them are presumably in contact with evolutionary theory. Finally, we hypothesized that (3) those (supposedly few) deviations from naturalist evolution may be explained by the educational environment of students (i.e. institutions with lower-quality standards).

We interviewed 1719 students from 123 graduate courses and other 123 postgraduate courses. Overall, the most common opinions were NaE (44.4%) and DGE (43.3%), followed by Cr (12.3%). The frequency of opinions did not differ according to area of interest (χ ²  = 49.32, df = 48, P = 0.42) or any possible interaction term (area*level*macroregion: χ ²  = 11.25, df = 16, P = 0.79; area*level: χ ²  = 13.42, df = 20, P = 0.86; area*macroregion: χ ²  = 24.50, df = 32, P = 0.82; level*macroregion: χ ²  = 29.64, df = 24, P = 0.19). However, we found that opinions differed according to student level (χ ²  = 76.28, df = 46, P < 0.001). Postgraduate students (Cr: 7.7; DGE: 38.2; NaE: 54.1%) were more prone to NaE at the expense of Cr and DGE when compared to undergraduate students (Cr: 14%; DGE: 47%; NaE: 39%; Fig. 1). Ultimately, opinions frequency differed according to geographical macroregion (χ ²  = 132.80, df = 52, P = 0.003; Fig. 2a), in which Cr and DGE increased northwards at the expense of NaE, which reached its peak in the regions further south.

Further analyses confirm this, with a significant difference according to geo-economic regions (χ ²  = 104.74, df = 4, P < 0.001). Amazonia did not differ from the Northeast (χ ²  = 7.3, df = 2, P = 0.027). On the other hand, the Center-South differed from both Amazônia (χ ²  = 42.41, df = 2, P < 0.001) and the Northeast (χ ²  = 85, df = 2, P < 0.001; Fig. 2b). Whereas DGE opinions remained similar in all three geo-economic regions, NaE was higher and Cr was lower in the Center-South.

Taking NaE as reference level, the multinomial logistic regression model was significant (χ ²  = 14.31, P = 0.026). Age and period do not have any effect on the probability of choosing Cr or DGE (Table 1). On the other hand, as the institution score decreased, Cr became more likely (Fig. 3a). The same pattern was marginally significant for DGE. When Cr and DGE are taken together, the opinions related to divine influence as a whole decrease as the score increase (χ ²  = 8.37, z = − 2.88, error = 0.38, P = 0.003; Fig. 3b).

Factors	DGE			Cr
	Error	T	P	Error	T	P
Age	0.025	− 1.060	0.288	0.032	0.321	0.747
Period	0.038	0.169	0.865	0.052	− 1.887	0.059
Score	0.607	− 1.765	0.077	0.805	− 2.937	0.003*

*Significant results at 0.05 level

The institution score proportion differed according to geo-economic region (F = 5.1, df = 2, P = 0.007). Amazonia (mean ± SD: 0.55 ± 0.14) and the Northeast (0.57 ± 0.12) presented lower scores compared to the Center-South (0.62 ± 0.14) (P = 0.03 and 0.049, respectively) and were not different from each other (P = 0.745; Fig. 3c). The score proportion did not differ according to level (F = 0.09, df = 1, P = 0.755) or the interaction term (geo-economic*level: F = 0.49, df = 2, P = 0.61).

Although biology students’ acceptance of evolution may increase gradually through the years of college (Paz-y-Miño and Espinosa 2009; Rocha et al. 2016), we did not find any changes, as in other studies (e.g. Bishop and Anderson 1990; Lawson and Worsnop 1992; Moore and Cotner 2009; Rice et al. 2011). This indicates that exposure to subjects with evolutionary tenor has no effect in students’ opinions during the graduate course. However, the evolution proportion increased among postgraduates. Based on the lack of increase during the graduate continuum, we suggest that the postgraduate selective processes gather students whose quality is related to evolution (despite elevated DGE and Cr views even among postgraduates).

When compared to Brazilian citizens, biology students are more biased toward NaE and less to Cr. The population in the country is only 8% NaE, followed by 59% DGE and 25% Cr (Datafolha 2010). However, assuming that Cr and DGE are both forms of divinely guided processes, these together comprise more than half of biology students’ opinions (55.6%). Paradoxically, this indicates that NaE is not the predominant explanation in students’ views. Nevertheless, Dias et al. (2012) interviewed students from the first graduation period from USP (University of São Paulo, Brazil), the largest university in South America, and found that the proportions of NaE, DGE, and Cr were 79.2, 19.8, and 1%, respectively. The authors compared these findings with those from the veterinary school of the same institution (49.1, 45.6, and 5.3%) and suggested that biology students are more prone to NaE even in early periods. Our results indicate that the national scenario is different from that of the USP first-period biology school. In fact, we found more Cr than Dias et al. (2012) did in the veterinary school, a course without contact with the theory of evolution. Why are the national parameters of biology students so dissimilar from those from USP (in biology and veterinary)? As evolutionary thinking is influenced by scientific literacy, USP probably has high-quality students already in the early periods. Our results concerning university score corroborate this, with students coming from low-scored institutions more likely to be inclined to Cr and divinely guided processes.

Based on our national parameters, we argue that the proportion of NaE among biology students is still low. Unfortunately, studies on evolutionary views are scarce in Brazil, and we still do not know which factors drive such a small proportion. For sure, problems concerning the quality of education in college and prior levels may influence evolution acceptance (Sinclair et al. 1997; Moore and Cotner 2009; Moore et al. 2011), but then again this information is lacking in the country. However, our unexpected marked regional differences may provide a clue, which is a fragile issue to discuss. It is known that 15-year-old students from states within the less developed macroregions (North and Northeast, most of them within Amazonia and Northeast geo-economic regions) present lower performance, probably due to the socioeconomic factor (INEP 2016b). In addition, when entering college, these students have access to low-scored institutions, as pointed out in our results. Thus, lower socioeconomic levels and institution quality are apparently related to reduced NaE acceptance among biology students. In fact, the NaE proportion found in the whole country may be related to the current weak status of education. Accordingly, among the 65 countries evaluated by the PISA (Programme for International Student Assessment), Brazil ranks 59th in science literacy, with low performance also in other areas (PISA 2012).

Finally, as “creationism is a symptom of religion” (Coyne 2012), the high proportion of divinely guided processes found may be a by-product of the intense religiosity of Brazilians. Issues related to personal beliefs including life, death, and origins may lead college students to respond more emotionally than rationally (Lawson and Weser 1990), and religion plays a key role in determining such beliefs (Sinclair et al. 1997; Alters and Nelson 2002). In general, people acquire their faith before having contact with evolution (Coyne 2012), most of them during childhood, being subjected to childhood indoctrination (sensu Dawkins 2006). Hence, even a graduate course deeply rooted in NaE may have no effect on extremely religious people. In the last national demographic census, 91.9% of Brazilians declared to have a religion (with no marked differences among the five macroregions) (IBGE 2012). Since adults traditionally transfer their values to the young (IBGE 2012), a considerable number of children grow up with a religion already determined. As a proxy of childhood indoctrination, for instance, 92.2% of children between 5 and 9 years old declare themselves to have a religion (IBGE 2012), long before having contact with any evolutionary theory.

In this study, we found that the majority of biology students paradoxically do not have NaE as an explanation, despite having presumed contact with the theory. In Brazil, opinions towards NaE were determined by socioeconomic and institution quality factors. We suggest that other attributes may also be important, such as low scientific literacy and the extreme religiosity of the population.

It is known that teachers naturally transmit their personal views to students (Carlsen 1991; Rocha et al. 2016). In addition, less economically developed countries present teachers with higher creationist conceptions (Clément 2015). Once biology students become researchers and teachers, our concern is that this NaE paucity may have a negative effect in classrooms in the long term, preventing the subject from being properly taught. Evolution is a fundamental issue in scientific literacy, providing groundwork for understanding a wide variety of topics in biology and other fields. Our study points out that evolution education needs to be reviewed in Brazil, and probably in other underdeveloped and emerging countries. We need a curricular/pedagogical reform aiming to fortify education on this important issue, from basic education to college.