INTRODUCTION

The commitment to research is one of the main characteristics of universities as academic training centers, present in their primary purposes and functions. Research is understood as an essential and mandatory function: "responding by producing knowledge and developing technologies to the needs of society, with special emphasis on the national reality" (^{Law No. 30220, 2014}).

Research work in higher education involves scientific and technical activities to promote science and innovation for the sake of social progress, and has three components: teaching, scientific and technical training, research and development, and scientific and technological services (^{OECD, 2002}). ^{Duarte (2013}) mentions that the key factors for research in a country are a good education, academic training, and the early start of scientific training.

For universities to fulfill this function, they need to have institutional capacities to create policies, strategies, and processes that support research practices sustainably and permanently with the support of the main actors: teachers and students. These and other elements are part of the requirements for the accreditation of universities in the country under charge of the National System of Evaluation, Accreditation, and Certification of Educational Quality (^{SINEACE, 2019}).

Various evidence of the difficulties students face in acquiring a research attitude is found, such as the technical report presented by the National Institute of Statistics (^{INEI, 2015}). This report reveals that of the total number of female graduates (women) from public and private universities in Peru, only 49.7% earned a professional degree. Meanwhile, of the total number of male graduates (men), only 36.4% earned their professional degree. Of all graduates (men and women) from public universities, 40.2% earned a professional degree, of which 53.4% opted for the thesis modality.

Regarding the positive perception of university graduates about the development of competencies, including those related to research, the following is evident: critical thinking, 60.5% in public universities and 60.3% in private universities; teamwork, 59.6% in public universities and 58.1% in private universities; report or document writing, 55.7% in public universities and 56.6% in private; presentation of ideas, 51.4% in public universities and 54.4% in private universities; use of computer tools, 49.1% in public universities and 56.2% in private universities, among others (^{INEI, 2015}).

The above results show the conditions of research teaching/learning at university, such as the inclusion of special subjects in which science and its methods are barely discussed, often only at the end of the degree program, and teachers who have not developed positive attitudes towards research (^{Olivera, 2020}). These aspects are reflected in the attitudes of students who are not willing to do research, even though it is a component of professional training and is applied in the labor market.

Likewise, the lack of motivation to research may be related to how reality is perceived and the lack of interest in changing it and/or promoting successful experiences. It is understood then that research not only contributes to training for professional performance, but also goes beyond this; science is progress and well-being and produces significant changes in societies (^{Piclín, 2008}).

^{Blanco (2008}) states that among the factors related to student research capabilities and competencies are attitudes which influence teaching/learning and immediate academic achievement. For ^{Allport (1935}), an attitude is a mental and neuronal state which is structured in experience and exerts a directive or dynamic influence in people’s response to the objects and situations with which they relate, giving rise to pleasant or unpleasant feelings and thoughts towards them, approving or disapproving them (^{Alonso et al., 2015}). For ^{Escalante et al. (2012}), they are theoretical constructs inferred from certain external behaviors, generally verbal.

Thus, attitudes are shown as a condition for research and as part of the qualities intended to be developed in university education. Thus, it is up to universities to foster, maintain and develop them within the framework of the so-called formative research. These, together with habits, skills, and knowledge, favor students to build knowledge and appropriate disciplinary concepts and theories to interact in the scientific world (^{Castro, 2017}); they also usually guarantee students’ motivation, interest, and effort (^{Quezada et al., 2019}).

The context plays a relevant role insofar as attitudes are acquired from other people through social learning, in situations in which one interacts with and observes their behavior; learning that occurs through various processes within the cultural and social environment (^{Baron & Byrne, 2005}). Attitudes are not expressed independently of culture, values, affections, experiences, or consciousness (^{Piclín, 2008}).

Likewise, ^{Kerlinger (2002}) defines scientific research as a systematic, controlled, empirical activity of rigorous approach to reality to know it, describe it, explain it, and/or understand it to transform it, if necessary. It implies assuming an onto-epistemic and methodological perspective when building knowledge (^{Ruiz, 2011}). For this study, the attitudes towards research construct is defined as an intentional, constant, and acquired process related to this activity, supported by one's values and perceptions.

Thus, knowing the student research attitudes is a previous step to implementing strategies to consolidate positive attitudes that contribute to fostering scientific research. According to the explanatory models, the most frequently used method for measuring attitudes is the Likert-type scale, which consists of a set of positive and negative statements related to a specific object; this type of scale is used to measure primary data, in which it is necessary to consider the social context in which the research is conducted (^{Blanco, 2011}).

To provide useful and reliable tools to account for attitudes towards research in specific contexts and thus provide coping strategies, measurement instruments have been constructed and/or validated, such as the Attitudes Towards Research Scale Revised Version (EACIN-R) by ^{Aldana et al. (2020}), which has been cited on several occasions and applied in countries such as Colombia, Mexico and Peru (^{Bendezú, 2021}; ^{Loayza-Rivas, 2021}; ^{Rojas-Solís et al., 2021}).

RESULTS

A total of 220 students in their final years of study (fourth and fifth) of the Anthropology, Archaeology, Geography, History, Sociology, and Social Work degree programs at the School of Social Sciences participated, of which 70 (31.8%) entered in 2016, 63 (28.6%) in 2017 and 87 (39.5%) in 2018. Of the total number of participants, 138 (62.7%) are women, the most representative group, with ages ranging from 20 to 32 years, 23 on average, and a standard deviation of 2.09 years. The male participants were 82 (37.3%), with ages ranging from 20 to 41 years, 24 on average, and a standard deviation of 3.72 years. The mean age of all participants was 23 years, with a standard deviation of 2.81 years.

When analyzing the psychometric properties of the Attitudes Towards Research Scale Revised Version (EACIN-R) in a Peruvian sample, Cronbach's alpha index was .898 for the test total, indicating good reliability. The reliability for each of its subscales was .862 for vocation, .692 for assessment, and .757 for research interest.

The total corrected correlation yielded values above .30 in most cases (24 items). Even though the correlation was lower than .30 in items 12, 19, and 21, it was close to this value; therefore, it is considered acceptable. The same did not occur with item 2, which yielded a value of .11, which is why special attention was paid to the quantitative analysis of this item, and the case was analyzed qualitatively by the research team to establish its relevance in the test. It was concluded that its inclusion was justified, following the researchers' experience and what is indicated in the literature.

Continuing with the analysis of the psychometric properties of the scale, an exploratory factor analysis (EFA) was made to examine the behavior of the items within the scale. Using the principal components method, a KMO of .792 was obtained, which is acceptable. MSA norm scores were also found for the items with a minimum value of .520 and a maximum value of .900. The values for each item exceeded .50, so it was decided to keep them all. With the principal components method and Promax rotation with Kaiser normalization, a value of .876 was obtained, similar to that obtained with KMO, confirming that the factor analysis is very good.

Likewise, Horn's parallel analysis was applied to confirm the multidimensional model (trifactorial) and the number of items of each subscale and, therefore, the total number of items of the instrument.

Subsequently, the multidimensional (trifactorial) model was confirmed with confirmatory factor analysis (CFA) to verify the theory underlying the scale. A graphic representation of the structure of the trifactorial model (path diagram) was obtained, showing the observable variables and the latent variables: research vocation, interest, and assessment (Figure 1).

Note. Adapted from ^{Aldana et al. (2020}).

Figure 1 Path Diagram of the model for measuring attitude towards research (EACIN - R) in a Peruvian sample. 

Next the model was identified, and it was found to have a unique solution. The parameters of the model were estimated with the Free-Scale Least Squares Method, suitable for ordinal scale variables. It should be noted that there is a wide variety of estimation methods (Maximum Likelihood Method, Weighted Least Squares Method, Unweighted Least Squares Method, Generalized Least Squares Methods, among others); in this particular case, the Free-Scale Least Squares Method suitable for ordinal scale variables was used (^{Jöreskog et al., 2016}).

It was verified that the structure of the trifactorial measurement model has a good fit. ^{Catena et al. (2003}), ^{Uriel and Aldas (2005}), and ^{Hair et al (1999}) suggest that the fit indices should be greater than .90 and that the RMR should be less than or equal to 0.08. The absolute goodness-of-fit statistic was a GFI of .942 > .90, AGFI of .932 > .932, and an RMR of .058 < .08, as well as the incremental goodness-of-fit statistics NFI of .922 > .90, RFI of .915 > .90, values that evidence good model fit (Table 1).

Table 1 Goodness-of-Fit Indices 

Source. Elaborated by the author

Likewise, high correlations between the latent variables were found, which, in all cases, are higher than .70 (Table 2).

Table 2 Relationship between the factors of attitude towards research. 

Source. Elaborated by the author.

DISCUSSION

Among the reasons why the psychometric properties of the EACIN-R scale were evaluated, is that as it is composed, with the subscales research vocation, assessment, and interest, with their respective items, it allows measuring intrinsic aspects of the subject, ranging from the inner call to research activity, to positions of discouragement that make it difficult to get involved in the research process. It is also possible to measure, from the scale, the recognition of the value of research for the advancement of knowledge and problem solution. The EACIN-R does not consider external elements that may influence attitudes, such as institutional conditions or academic load.

The aim was to confirm the psychometric properties of the Research Attitudes Scale-Revised Version (EACIN-R) in Peruvian university students. Good internal consistency was obtained in that Cronbach's alpha index for the test total was .898 for the subscale research vocation, the value was .862 for research assessment .692, and research interest was .757. The correlation was positive and statistically significant among the three subscales.

Horn's parallel analysis confirmed the multidimensional model (trifactorial), the number of items corresponding to each subscale, and the total number of items of the instrument, showing that the EACIN-R is a useful and reliable tool to measure this construct in Peruvian students, thus making it possible to contribute to identifying factors to promote research activity in the country.

The results found are consistent with the analysis of other instruments in which similar measures were found for internal consistency, such as the scale reported by ^{Muñoz et al. (2020}), which yielded a value of .91 with the Omega coefficient and which contains subscales called intrinsic student motivations, training quality and institutional context. Likewise, they are similar to what was found by ^{Castro (2017}) in the design and validation process of a scale to evaluate attitudes towards formative research, which obtained the value of .827 with Cronbach's alpha index. The instrument is composed of five factors: perceived abilities, learning behaviors, satisfaction/agreeableness, conceptual appropriation, and systematic exploration. However, it is focused on formative research, not research in general, which is the purpose of this study.

On the contrary, our results differ from the analysis of other instruments, such as the one validated by ^{Ramos (2019}) that yielded a Cronbach's alpha of .54 and the one validated by ^{Arellano-Sacramento et al. (2018}) with a Cronbach's alpha of .65, both of which indicate low internal consistency. These instruments are composed of dimensions such as the teacher as a model, proactivity, disinterest in science, rejection of research activities, the exclusivity of scientific activities (Ramos, 2019), attitude towards the production of scientific articles, attitude towards the actions of the university to encourage research in students, among others, (Arellano-Sacramento, 2018), dimensions different from those contained in the EACIN - R.

In terms of theoretical and practical implications, the results obtained indicate that the EACIN-R scale is reliable and useful to measure attitudes towards research in Peruvian university students. It is recognized that one of the weaknesses of this study was not having had a larger sample size due to the COVID-19 pandemic since the number of items is 28, which merited at least 280 participants to meet the quota of at least 10 participants per item (^{Kline, 2005}).

CONCLUSIONS

The confirmation of the psychometric properties of the EACIN-R scale provides an instrument suitable for application in a population with similar characteristics to the one that participated in this study. Thus, it contributes with a useful tool to identify attitudes towards research in university students and eventually design guidelines to strengthen research activity in the country. Its use is feasible in different areas of knowledge to promote research, considering that it is one of the main functions of higher education and that attitudes towards it play an important role.

It is ratified that the EACIN-R is useful to measure attitudes towards research in Peruvian university students because its multifactorial structure of three factors: vocation, interest, and valuation, following the underlying theory, was confirmed. In addition, Cronbach's alpha was .898 in the test total, meaning a good internal consistency.

It is suggested to investigate attitudes towards research in relation to other variables, such as sociodemographic contexts, comparative analysis between groups, gender perspective, or other aspects considered of interest. Following the recommendations of the specialized literature, it is important to include a larger sample size so that there are at least 10 participants for each item.

RECOMMENDATIONS

For future research, it is recommended to evaluate the concurrent criterion validity of the EACIN-R in different educational and cultural contexts