Research methods anxiety, attitude, self-efficacy and academic effort: A social cognitive theory perspective

Abstract

The integration and mandatory study of research method courses by university students is a global phenomenon. However, students demonstrate apprehension and disinterest in research methods courses. Consequently, this study investigated the interplay among research methods anxiety, positive disposition toward research, perception of usefulness, self-efficacy, and academic effort in research. Paper-and-pencil questionnaires were used to collect data from a convenience sample of 636 technical university students in Ghana. Based on partial least squares structural equation modeling results, research anxiety reduces students’ self-beliefs (β = −0.172; t = 2.729; p = .000) and negatively affects perception of research usefulness (β = −0.172; t = 2.729; p = .000). A positive disposition towards research enhances students’ academic efforts in research (β = 0.919; t = 3.308; p = 0.001). Furthermore, self-efficacy drives students’ efforts (β = −0.172; t = 2.729; p = .000), enhances positive disposition toward research (β = 0.191; t = 3.308; p = .001) and perception of research usefulness among students (β = 0.224; t = 3.064; p = .002). Students’ perception of research usefulness mediates the relationship between anxiety and effort in research (β = −0.075; t = 3.653; p = .000), positive research disposition and effort (β = −0.056; t = 2.519; p = .012) as well as research self-efficacy and effort in studies (β = 0.083; t = 3.02; p = .003).

Keywords:

• Research anxiety
• attitude
• self-efficacy
• usefulness
• efforts
• university students

1. Introduction

Anxiety is a debilitating mental health condition prevalent among college students across the world. The American Institute of Stress [AIS] (2019) estimates that about 75% of all university students in the US in 2017 experienced at least one episode of “overwhelming anxiety”. The American Psychological Association [APA] (2022) defined anxiety as an emotion characterized by feelings of tension and worried thoughts. Typical signs of anxiety include nervousness, unease, a sense of impending doom, sweating and trembling, inability to maintain focus, uncontrollable worry, insomnia, panic attacks, headaches, and high irritability (Gilchrist & Ditto, 2015; American Institute of Stress [AIS], 2019). Anxiety negatively affects students’ academic achievement (Jamil et al., 2022; Vitasari et al., 2010; Carey et al., 2016; Weda & Sakti, 2018) as they usually experience higher levels of anxiety in performance situations such as tests and examinations. Importantly, anxiety may induce poor performance and poor performance may induce anxiety (Carey et al., 2016; Maharajan et al., 2017).

As argued by Carey et al. (2016) however, the manifestation of anxiety is not restricted to tests and examinations alone; it also applies to different academic subjects in school. Anxiety in a school subject may be influenced by perception of difficulty of that subject or by prior experience of failure in the subject. It has been established on many occasions that a student’s test anxiety score is related to his/her perception of a subject’s difficulty, regardless of the subject or demands of the test in question; and that a student’s perception of a subject’s difficulty is positively correlated with that student’s anxiety levels in the subject (Carey et al., 2016). Although scholarly efforts at understanding academic subject anxiety have tended to focus preponderantly on statistics and mathematics, there is an emerging concern and some nascent evidence to suggest that research methods courses pose anxiety problems similar to those related to statistics (Ashrafi-Rizi et al., 2015; E. C. Papanastasiou, 2005; Gurung & Stoa, 2020; Onwuegbuzie, 2004; Papanastasiou & Zembylas, 2008) among undergraduate students. Interestingly there is very limited empirical examination of the issue of research methods anxiety among university students, globally (Papanastasiou & Zembylas, 2008). Of particular concern is the dearth of empirical work on the issue in developing regions, especially Africa, as an overwhelming proportion of current scholarship on the subject have tended to focus on university students in the developed world. Such neglect has served to restrict current knowledge on the issue; stymie effective theory development on it; and undermine efforts at developing credible practical interventions for addressing the challenge among university students, more broadly. By focusing on Ghana, this study attempts to address this obvious neglect by providing empirical evidence from the African region that has so far been largely overlooked in current scholarship on the subject.

Generally, it is long-established that research methods courses constitute an important part of the curriculum in almost all undergraduate programs to help students improve their skills in general enquiry and critical thinking; acquire the basic skills needed to complete their own research and gain subsidiary skills necessary for their professional life (Daniel, 2018; Petrella & Jung, 2008). In line with these aims, research methods courses have become a key requirement in almost all undergraduate programs, and all students must take a course in research methods in order to graduate (Papanastasiou & Zembylas, 2008; Jansen et al., 2021). Studies have however confirmed that students tend to view research methods courses negatively and experience anxiety towards them (Papanastasiou, 2005; Papanastasiou & Zembylas, 2008; Einbinder, 2014; Gredig & Bartelsen-Raemy, 2018). Research methods anxiety—the anxiety experienced by students when they are confronted with research methods content—is found to be significantly related to four main factors: attitude, research usefulness, self-efficacy, and academic effort (Ashrafi-Rizi et al., 2015; Iovu et al., 2015; Li, 2012; Maharajan et al., 2017).

Attitude is the negative or positive feeling of a person arising from experience or implicit learning, which may regulate the person’s response to a thing, subject or others (Hammad, 2016). According to Schutt et al. (1984), some students see research methods as a “field of boulders” and take it only because it is required. Research usefulness indicates a student’s perception of the value of research methods course for their program of study or future professional career (Hu et al., 2009; Lai & Wang, 2012). Papanastasiou and Zembylas (2008) suggest that most undergraduate students do not recognize the usefulness of research methods courses for their careers and may not understand why they need to take that course. Self-efficacy is a person’s judgment of his or her abilities about a particular issue or subject (Ashrafi-Rizi et al., 2015; Tahmassian & Jalali Moghadam, 2011). Low levels of self-efficacy predispose one to high levels of anxiety with attendant consequences for academic achievement (Tahmassian & Jalali Moghadam, 2011). Academic effort addresses the amount of time and energy committed by a student in meeting the requirements of an academic subject (Carbonaro, 2005).

Recent studies exploring the linkages among attitude, research usefulness, self-efficacy, academic effort, and anxiety reveal some interesting insights. Iovu et al. (2015) have found a positive correlation between attitude, self-efficacy, and academic effort, with attitude and self-efficacy serving as significant predictors of effort. Another study by Maharajan et al. (2017) in Kuala Lumpur also shows that academic effort and self-efficacy are negatively correlated with students’ research anxiety. It is further found that high perceptions of research usefulness may lead to positive attitude but may not necessarily engender higher academic effort toward research methods courses . The link between attitude and research methods anxiety is less certain. Many studies have established the correlation between attitude and anxiety; most have also found attitude as a strong predictor of academic achievement (Li, 2012; Oguan Jr., Bernal, & Pinca, 2014) although some questions remain concerning the causal links between the two variables (Iovu et al., 2015). Yet, according to Morgenshaten et al. (2011), positive attitudes to research may not necessarily reduce students’ anxiety towards the subject (Morgenshaten et al., 2011).

These findings raise some uncertainties around the linkages between and among these variables that warrant further investigation in order to facilitate useful remedial innovations in addressing research methods anxiety. Of particular interest is the role of attitude in anxiety and its interrelationships with other variables such as academic effort, self-efficacy and research usefulness. Indeed, as Wisecup (2017) argued, there is a gap in literature relative to students’ attitudes to research methods courses as much of what is known currently in this regard is only anecdotal and requires more empirical scrutiny. Importantly, current research findings indicating that positive attitude may not necessarily reduce research methods anxiety necessitates further studies to shed light on the interactions among the key constructs under study: attitude, research usefulness, self-efficacy, academic effort.

The purpose of this study is to investigate research methods anxiety among undergraduate students in a Ghanaian technical university. Significantly, by focusing on Ghana, the study provides a new regional perspective to the discussions on anxiety among undergraduate students and fills an important void in the current literature on the subject. The study focuses on not just documenting the level of prevalence of research methods anxiety among the students, which is itself extremely important given the dearth of such studies in Ghana, specifically; it also focuses on exploring and shedding light on the interlinkages among academic effort, attitude towards research, self-efficacy and research anxiety (Figure 1) in the context of the study of research methods courses among Ghanaian university students. Specifically, the study focuses on the following objectives:

1. Assess the influence of research anxiety on students’ positive research disposition, perception of research usefulness, self-efficacy and academic effort in research.

2. Evaluate the effect of positive research attitude on perception of research usefulness and research academic effort of students.

3. Explore the relationship between self-efficacy and positive research disposition, perception of research usefulness and academic effort in research

4. Examine the mediating role of research usefulness between anxiety, positive attitude, self-efficacy and academic effort in research.

Figure 1. Conceptual Framework and proposed hypotheses.

The study makes the following contributions to the existing literature. First, the study examined research method anxiety as an antecedent of four main concepts including academic efforts, self-efficacy, research usefulness and research attitude. The findings of the study validate earlier works that have reported the effects of anxiety on students’ attitudes from different subject context (Chetri et al., 2022; Liaw & Huang, 2013; Schlebusch, 2018; Wishkoski et al., 2022). Second, the study further offers evidence to support the mediating role of research usefulness on the relationship between (i) research method anxiety, (ii) attitude, and (iii) self-efficacy on academic effort, a relationship that remains unexplored.

2. Literature review and hypothesis development

2.1. Theoretical framework

Social cognitive theory (SCT) and expectancy-value theory (EVT) provide the theoretical underpinnings for the current study. SCT is a human behavior theory based on the work of Bandura, 2010). The social cognitive theory incorporates both behavioral and cognitive perspectives to explain human behavior. Based on the theory, three factors influence human behavior and these are the person, behavior and environmental factors. Importantly, these influencers affect and cause each other, giving rise to the concept of triadic reciprocal determinism. One important component of the theory is the construct of self-efficacy. Self-efficacy highlights the self-beliefs of individuals about their ability and competency to accomplish a task. Self-efficacy is determined by many factors such as emotions (anxiety) and some of the outcomes of self-efficacy include persistence on a task, effort as well as achievement. In the context of the current study, students’ research self-efficacy is likely to be influenced by academic effort in research methods and vice versa. Furthermore, self-efficacy beliefs insulate students from symptoms of cognitive anxiety. In this case, lower levels of self-efficacy will trigger higher levels of anxiety.

The expectancy-value theory (ETV) is an important behavioral theory that has been applied in several fields including learning motivation within the educational setup. The theory posits that human behavior and actions are explained by expectancy regarding the success on an activity as well as the value that is placed on the activity (Wigfield & Eccles, 2002). Expectancy is about the conviction of being successful at executing a task and this perception of success drives effort, choice of tasks, persistence as well as achievement. The value component of the theory is divided into four: intrinsic or interest; attainment value, utility value and cost. The intrinsic or interest value is about the anticipated enjoyment one expects to gain from undertaking a learning activity whereas attainment value or importance refers to whether success on a task will support or confirm the student’s valued characteristics. The utility value or usefulness is about students’ perception regarding the significance of research learning activity contributing to the achievement of present or future plans. Lastly, cost refers to what students have to forego in order to do a task as well as the expected effort one will need to put into task completion (Wigfield & Eccles, 1992, 2000). Based on the perspective of expectancy-value theory, it is expected that students’ self-beliefs in research will determine the amount of effort they expend on research related learning activities. Furthermore, students’ perception of usefulness of research to their future career goals will determine the academic efforts student put into research learning activities.

2.2. Anxiety and research academic effort

Research anxiety is a complex array of emotional reactions, which occur when a student encounters research methods in any form and at any level (Papanastasiou & Zembylas, 2008). Anxiety interferes with students’ concentration and memory thereby affecting their academic work negatively (Onwuegbuzie & Seaman, 1995). Meanwhile, the efforts of students on a course positively correlate with academic achievement (Stewart, 2008). In an Australian study conducted among 147 selected undergraduate students enrolled in psychology and social sciences programmes, statistics anxiety was related to academic effort (Macher et al., 2011, 2013) as students with high statistics anxiety were more likely to invest less effort and time for learning, use less efficient learning strategies. A Malaysian study reported moderately negative correlation between students’ academic effort towards research methods and research anxiety (Maharajan et al., 2017). Based on the above evidence, we hypothesize that,

H₁: There is a significant negative relationship between research anxiety and academic effort of students in research methods.

2.3. Anxiety and attitude

A positive attitude toward research enhances students’ educational success (Al-Kuwaiti, 2014). Students’ difficulties in research methods studies are reflections of attitudinal factors such as anxiety rather than insufficient aptitude (Balog˘lu, 2001). High levels of anxiety correlate negatively with positive attitudes in various subjects (Akin & Kurbanoglu, 2011; Cazan, Cocorada, & Maican, 2016; Korobili et al., 2010;). For example, a strong negative relationship was reported between students’ computer anxiety and positive attitude in a study conducted among students of Library and Information Systems from the Department of Technological Educational Institute of Thessaloniki (Korobili et al., 2010). In a sample of Romanian high school and university students, Cazan et al. (2016) found that students with a high level of anxiety also had a negative attitude towards the use of the internet. Similarly, among university students enrolled in Sakarya University in Turkey, positive math attitudes were negatively related to math anxiety (Akin & Kurbanoglu, 2011). Based on the foregoing, the following hypothesis was stated.

H₂: There is a significant negative relationship between research anxiety and students’ positive attitude towards research methods.

2.4. Anxiety and research usefulness

Students’ perception of research usefulness is rarely investigated in the context research of teaching and learning. Similarly, the influence of anxiety on perception of usefulness or task value has attracted limited empirical attention. Research usefulness is the degree to which a student perceives research methods as beneficial to the attainment of future goals (Lai & Wang, 2012). However, few studies have investigated the relations between these concepts in the context of computer literature, and the conclusions of these studies are inconclusive. Within the context of mobile learning adoption, Callum et al. (2014) found a negative relationship between ICT anxiety and perceived usefulness of mobile learning. However, in the study of Liaw and Huang (2013), the relationship between perceived anxiety and usefulness of e learning was unsupported. On the premise of the following, it was hypothesized that:

H₃: Research anxiety will influence perception of research usefulness

2.5. Anxiety and self-efficacy

Self-efficacy is an important personal concept in social cognitive theory that predicts students’ performance outcomes (Schunk et al., 2008; Usher & Pajares, 2008) and is strongly related to motivation and cognition as it influences students’ task performance, interest and persistence, goal setting, and the choices they make (Lent et al., 2002; Linnenbrink & Pintrich, 2003; Schunk, 2003). Self-efficacy beliefs protect students from cognitive and affective anxiety symptoms . Consequently, higher levels of self-efficacy lead to lower anxiety levels albeit inconclusiveness. Students with higher level of self-efficacy report lower levels of anxiety towards a task (Munoz et al., 2018; Qudsyia & Putrib, 2016; Tahmassian & Jalali Moghadam, 2011). However, some studies have reported either weak or no correlation between these variables. For example, in a study investigating the relationship between employees’ computer anxiety and self-efficacy in Malaysia Civil Defense Head Quarters, Kuala Lumpur, a weak correlation was found between the two concepts (Achima & Al Kassimb, 2015) while others found insignificant relationship between students’ anxiety and self-efficacy (Durndella & Haagb, 2002; Mills et al., 2006). Based on the above evidence, we hypothesized that:

H₄: There is an inverse relationship between research anxiety and student’s research self-efficacy.

2.6. Attitude and academic effort

Even though studies on attitude and academic effort are rare, few studies that examined the relationship between attitude and effort have reported significant positive outcomes. On the assumption that attitudes influence behavior, it is expected that when students show positive attitudes towards a subject, they are likely to expend more time in learning activities related to the subject. For example, Li (2012) studied 153 students at City University of Hong Kong and found that attitude positively predicted academic effort. In another study, Iovu et al. (2015) found attitude significantly predicted the effort of 109 social work students towards research methods. In an Australian study involving 56 respondents, Hemmings and Kay (2010) reported a significantly positive association between mathematics attitude and amount of effort students expend in the subject. On the premise of the foregoing, we hypothesize that:

H₅: There is a significant positive relationship between attitude and academic effort of students in research methods.

2.7. Attitude and usefulness

Evidence from ICT related studies suggests experience is an important influencer in the way people perceive a new technology as being useful, and affects their attitude towards using it (Lymperopoulos & Chaniotakis, 2005; Poon, 2008). Jahangir & Begum, 2007) studied 227 electronic banking users of private commercial banks in Bangladesh and found that perceived usefulness of new platform had positive effect on customer attitude towards the introduction of electronic banking. In another study, Selevičienė and Burkšaitienė (2015) found that perception of usefulness of web 2.0 was associated with a positive attitude of 101 students of Mykolas Romeris University in Lithuania. Guritno and Siringoringo (2013) reported a strong link between perceived usefulness and attitudes towards usability of airlines ticket reservation among 283 clients. Based on the above reviewed evidences, we hypothesize that:

H₆: There is a significant positive relationship between research usefulness and attitude of students towards research methods.

2.8. Research usefulness and academic effort

Based on the expectancy-value theory, perception of research usefulness will relate to the effort students put in learning research methods. The expectancy-value theory suggests that students’ motivation to participate in learning activities is dependent on the value placed on the task (Wigfield & Eccles, 1992, 2000). In this case, when students perceive research methods as relevant to the achievement of future goals, they are likely to invest a lot of effort in learning activities related to the course. The expectancy-value theory has received empirical support in the context of the higher education setup (Cole et al., 2008; Dietrich et al., 2017; Schmid & Bogner, 2015). In a quasi-experimental design study of high-achieving 9^th graders, a positive significant relationship was reported between effort and perception of usefulness (Schmid & Bogner, 2015). In a study within the context of assessment, usefulness predicted test-taking effort of undergraduate students in the USA (Cole et al., 2008). Based on the foregoing, it was hypothesized that:

H7: There is positive relationship between research usefulness and research effort

2.9. Self-efficacy and academic effort

According to Dissanayake et al. (2019), self-efficacy influences the amount of effort students allot and the continuity of that effort. A review of extant literature suggests that limited empirical attention has been paid to the relationship between self-efficacy and academic effort within the domain of research teaching and learning. However, a study conducted in Spain among undergraduate students indicated that students’ higher self-efficacy levels were related to more effort in academic studies (Valle et al., 2009). Within the context of competitive crowdsourcing funding platform for academic, self-efficacy positively affected effort (Dissanayake et al., 2019). Lastly, in a study of undergraduate engineering students in a major research university in the southwest, US, Wu, Fan, Weihua, Arbona, and Rosa-Pohl (2020) found that self-efficacy was a positive significant contributor to students’ academic effort. On the premise of the foregoing, it was hypothesized that:

H₈: There is a significant positive relationship between students’ perceived self-efficacy and academic effort in research.

2.10. Self-efficacy and attitude

Several studies have reported a positive relationship between self-efficacy and positive dispositions of students. Natividad et al. (2019) examined 154 third year students and found a strong association between self-efficacy and attitude towards research. In another study conducted among 784 students in India, Kundu and Ghose (2016) found students’ attitude towards the study of mathematics to strongly influenced their self-efficacy. Finally, Yau and Leung (2018) researched 187 students in a Hong Kong higher education learning institution and found positive and significant relationship between attitudes towards the use of technology and their self-efficacy. Based on the above evidence, we inferred the following hypothesis:

H₉: There is a significant positive relationship between self-efficacy and students’ positive attitude toward research methods.

2.11. Self-efficacy and usefulness

Results of several studies have suggested that self-efficacy influences perception of usefulness within educational context, and most of these studies investigating the relationship between these concepts are within the domain of information technology (Liaw & Huang, 2013; Muslichah, 2018). For example, a study investigating 194 university students’ attitudes toward iCAN e learning concluded that self-efficacy was a driver of perceived usefulness of iCAN e learning (Liaw & Huang, 2013). Another study in Nigeria involving pre-service teachers showed that self-efficacy influenced their perceived usefulness of information technology for teaching (Shittu et al., 2016). In the context of social networking sites adoption among undergraduate students in a leading business management university in Thailand, Surej (2013) found a positive effect of self-efficacy on perception of usefulness of social networking sites. Based on the foregoing empirical evidence, it was hypothesized that:

H₁₀: There is a significant positive relationship between self-efficacy and students’ perceived usefulness of research methods.

2.12. Mediating effect of research usefulness and academic effort

Based on the established relationships from the reviews conducted, we expect that students’ perception of research usefulness would further enhance the relationship between (i) anxiety and academic effort, (ii) attitude and research academic effort and (iii) self-efficacy and academic effort. Put differently, we expect usefulness of research to reduce student’s anxiety towards research thereby enabling them to enhance effort towards research. We therefore inferred the following hypotheses:

H₁₁: Research usefulness mediates the relationship between anxiety and academic effort

H₁₂: Research usefulness mediates the relationship between attitude and academic effort

H₁₃: Research usefulness mediates the relationship between self-efficacy and academic effort

3. Methods

The study population constituted students in a small-sized technical university located in Ho, the Volta Region of Ghana with student population estimated at 4500 as of the time of collecting data for the study. Students study a variety of programs ranging from engineering, business, computer science, food science to hospitality management. A greater proportion of the students study at the higher national diploma (HND) level while a minority pursue top-up degree programs. Students in the university are supposed to submit an independent and supervised project work/dissertation as a partial requirement for graduation. In order to equip students with research skills and competencies to write project works, all students take research methods courses usually during their second year of schooling. The study was conducted within the framework of a cross-sectional research design.

3.1. Sample and procedure

Given that the study sought to investigate anxiety and attitude of students toward research methods courses, only second and third year students were involved in the study because research methods courses are available in the second year of schooling. In determining the sample size for the study, Taro Yamane’s formula was used, n = N/(1+ Ne²), where n is the sample size, N is the population size and e is the precision level. The population of the target group was estimated at 3530 with estimated precision of 0.05 therefore; using the formula, the sample size was determined to be 359. However, to take care of non-responses and misplacement of the survey forms, 700 questionnaires were distributed proportionally among students across all the four academic faculties in the university. Out of 21 academic programmes in the university, seven were randomly selected to participate in the study. Quota sampling technique was used to allocate the number of respondents to select from each programme and study year. Available and willing students participated in the study.

Paper-and-pencil questionnaires were distributed in classrooms after the authors sought help from colleague lecturers to facilitate the data collection process. To ensure that the data collection procedures conform to ethical requirements, before handing over questionnaires to students, they were assured of voluntary participation, anonymity and confidentiality. To assure anonymity, the research participants were not required to provide their names or student registration numbers on the questionnaires. Students dropped completed questionnaires in boxes placed at the entrance of classrooms. This procedure ensured that students did not feel compelled to participate in the survey because their lecturers were leading the data collection process. All the questionnaires were retrieved, but 64 were discarded because of incompleteness, leaving 636 giving a response rate of 90.8%. Regarding sample characteristics, more than half of the respondents were female (56.3%) while males constituted 43.7% of the sample. Over half (57.8%) of the students were aged 21–25 years followed by those between 26 and 29 (20.5%) age group while those in the age brackets of 30 or more (11.4%) and 20 or less (10.2%) were in the minority. A majority of the respondents (75.5%) were third year (final year) students while second year students constituted 24.5% of the sample. Business and management students (accounting, banking and finance) constituted almost half (47.2%) of the sample followed by hospitality management (26.6%) and engineering (26.1%) students (mechanical, building technology, computer).

3.2. Measures

A revised version of the Attitudes toward Research (R-ATR) scale (E. Papanastasiou, 2014) consisting of three sub dimensions (research usefulness, anxiety and positive research attitude) was used to measure students’ attitude towards research methods. The Attitudes towards Research scale has been used in earlier studies investigating students’ attitude toward research methods courses (Jansen et al., 2022; Wishkoski et al., 2022). The Research Usefulness (RU) subscale consisted of nine items. Sample item was “research is useful for my career”. The Research Anxiety (RAN) subscale was measured with eight items. Sample item was “research makes me anxious”. The Positive Research Attitude (PRA) was measured using 10 items. Sample items included “I like research”. To measure Academic Effort in Research (AER) and Research Self-Efficacy (RSE), 15 items were adopted from Maharajan et al. (2017). Examples of the 7-item instrument that measured AER included “I work hard to complete research assignments”. In the case of RSE, an example of the 8-item instrument included “I am able to learn research methodology and data analysis”.

3.3. Data analytic procedure

The demographic profile of respondents was analyzed using descriptive statistics whereas the research hypotheses were assessed with PLS-SEM version 3.2.3 (Hair et al., 2017). In line with guidelines provided by Hair et al. (2017), the measurement model was assessed followed by the structural model. Additionally, Common-Method Variance (CMV) was tested using Bagozzi et al.’s (1991) method. As shown in Table 2, the highest correlation between any of two constructs is 0.683 < 0.90 (Bagozzi et al., 1991) (correlation between AER and RSE). Hence, no CMV was found in the data collected.

4. Results

4.1. Measurement model assessment

The results relating to outer model evaluation are set out in Tables 1 and 2. In order to establish convergent validity and reliability for the studied constructs, items with correlation coefficient values less than 0.7 were removed, and in this respect, two items were deleted in the case of PRA. As shown in Table 1, the Cronbach’s Alpha of AER, RAN, PRA, RU and RSE exceeded the recommended threshold of 0.70 (Nunnally, 1978) thereby confirming the reliability of all measures. Similarly, both average variance extracted and composite reliability values exceeded the permissible threshold of .50 and .70 respectively, thus affirming the reliability and validity of the model’s latent variables (Hair et al., 2017; Bagozzi & Yi, 1988).

Table 1. Validity and reliability of latent constructs

Construct	Cronbach’s Alpha	Composite Reliability	AVE
AER	0.876	0.910	0.670
RAN	0.894	0.913	0.568
PRA	0.892	0.916	0.611
RU	0.915	0.931	0.629
RSE	0.870	0.902	0.605

Table 2. Discriminant validity (Fornell-Larcker Criterion and Heterotrait-Monotrait Ratio (HTMT)

Fornell-Larcker Criterion						HTMT
	AER	RA	PRA	RU	RSE	AER	RA	PRA	RU	RSE
AER	0.819
RAN	−0.198	0.754				0.19
PRA	0.570	−0.086	0.781			0.643	0.151
RU	0.567	−0.375	0.450	0.793		0.630	0.389	0.496
RSE	0.683	−0.171	0.608	0.517	0.778	0.771	0.190	0.693	0.573

Three indicators were used to assess the discriminant validity of the measurement model. First, Fornell and Larcker (1981) procedure was applied. As shown in Table 2, the square root of the AVEs of each construct in the matrix diagonal is higher than the related correlation in corresponding rows and columns (Hair et al., 2017). Secondly, the Heterotrait-Monotrait proportion of relationships (HTMT) criteria for each pair of reflective constructs based on the item correlations (Henseler et al., 2015) was estimated (Table 2). The results from the correlations of pair of constructs are less than the threshold values of HTMT = 0.90 (Henseler et al., 2015).

Finally, the cross-loading values of the reflective construct indicators were also determined. As shown in Table 3, all indicators of reflective measurement models met the cross loadings assessment criteria (Hair et al., 2017). The results from the three methods demonstrate the quality of the reflective model and a confirmation of the model’s discriminant validity.

Table 3. Cross-loading among reflective measurement scale items

	AER	RAN	PRA	RU	RSE
AER 1	0.854	−0.169	0.503	0.523	0.645
AER 2	0.823	−0.124	0.420	0.450	0.565
AER 3	0.735	−0.160	0.412	0.364	0.536
AER 4	0.830	−0.188	0.518	0.474	0.547
AER 5	0.846	−0.170	0.472	0.498	0.493
PRA 2	0.411	0.030	0.822	0.328	0.476
PRA 3	0.476	−0.041	0.856	0.388	0.494
PRA 4	0.446	−0.134	0.850	0.421	0.521
PRA 5	0.423	−0.197	0.738	0.383	0.433
PRA 6	0.437	−0.048	0.743	0.317	0.428
PRA 7	0.451	−0.030	0.736	0.279	0.479
PRA 9	0.468	−0.046	0.711	0.337	0.487
RA 1	−0.087	0.729	0.069	−0.264	−0.062
RA 2	−0.136	0.803	−0.020	−0.288	−0.109
RA 3	−0.239	0.762	−0.099	−0.384	−0.202
RA 4	−0.097	0.782	0.016	−0.247	−0.060
RA 5	−0.008	0.709	0.060	−0.187	0.029
RA 6	−0.204	0.779	−0.118	−0.279	−0.184
RA 7	−0.137	0.746	−0.038	−0.216	−0.070
RA 8	−0.127	0.715	−0.211	−0.273	−0.186
RSE1	0.393	−0.039	0.529	0.282	0.748
RSE2	0.445	−0.059	0.525	0.357	0.767
RSE3	0.531	−0.153	0.501	0.445	0.795
RSE4	0.552	−0.176	0.428	0.374	0.802
RSE5	0.578	−0.160	0.419	0.438	0.766
RSE6	0.652	−0.188	0.454	0.487	0.790
RU 1	0.437	−0.272	0.406	0.792	0.415
RU 2	0.473	−0.318	0.361	0.819	0.411
RU 3	0.471	−0.342	0.393	0.826	0.379
RU 4	0.411	−0.208	0.310	0.765	0.410
RU 5	0.489	−0.323	0.406	0.864	0.440
RU 6	0.464	−0.321	0.334	0.805	0.423
RU 7	0.432	−0.340	0.306	0.760	0.411
RU 8	0.417	−0.244	0.333	0.705	0.395

4.2. Analysis of structural model

The standardized root means square residual (SRMR) value was used to assess the overall model fit (Henseler et al., 2016). For the study, the SRMR value was 0.072 < 0.08, representing a good model fit (Hu & Bentler, 1998). The Stone-Geisser’s Q² Test (Geisser, 1974; Stone, 1974) is used to evaluate the predictive accuracy of the exogenous latent variables in the model using blindfolding with an omission distance of 7. As shown in Table 4, apart from RAN, Q² values were significantly above zero, confirming the exogenous constructs’ high predictive relevance (Chin, 1998).

Table 4. Hypothesis and effect size assessments

Path	β	t-Values	p-Values	f^2(effect size)	Q^2	R-squared adjusted
RAN -> AER	−0.013	0.286	0.775	0.000
RAN-> PRA	0.019	0.324	0.324	0.001
RAN-> RU	−0.299	5.894	0.000	0.141
RAN-> RSE	−0.172	2.729	0.000	0.030
PRA-> AER	0.191	3.308	0.001	0.049
PRA -> RU	0.224	3.064	0.002	0.051
RU-> AER	0.251	4.567	0.000	0.087
RSE -> AER	0.436	6.348	0.000	0.235
RSE-> PRA	0.611	12.52	0.000	0.575
RSE->RU	0.330	4.599	0.000	0.109
RAN-> RU-> AER	−0.075	3.653	0.000
PRA-> RU-> AER	−0.056	2.519	0.012
RSE->RU->AER	0.083	3.02	0.003
AER					0.342	0.546
RAN					0.011	0.026
PRA					0.207	0.366
RU					0.223	0.378

The hypotheses were assessed based on the conventional significance level of 0.05. As shown in Table 4, against expectation, the hypothesized relationship between RAN and AER (β = −0.013; t = 0.286; p = 0.775) and RAN and PRA (β = 0.097; t = 1.485; p = 0.138) were insignificant thereby rejecting hypotheses H₁ and H₂. The estimated results indicate a negative significant relation between RAN and RU (β = −0.299; t = 5.894; p = .000), lending support for hypothesis H₃. The expected negative relationship between RSE and RAN was supported (β = −0.172; t = 2.729; p = .000). In support of hypotheses H₅ and H₆, the path relationships between PRA and AER (β = 0.191; t = 3.308; p = .001), and PRA and RU (β = 0.224; t = 3.064; p = .002) were significantly positive. Similarly, the path relation between RU and AER (β = 0.251; t = 4.567; p = .000) was significant thereby providing support for H₇. As expected, the hypothesized positive relationship between RSE and AER was supported (β = 0.436; t = 6.348; p = .000). RSE was significantly and positively related to both PRA (β = 0.611; t = 12.52; p = .000 and RU (β = 0.330; t = 4.599; p = .000) thereby supporting hypotheses H₉ and H₁₀ respectively. Regarding mediation effects, in support of hypotheses H₁₁, H₁₂, H_13, results of the study show that RU fully and partially mediates the relationship between RAN and AER (β = −0.075; t = 3.653; p = .000), PRA and AER (β = −0.056; t = 2.519; p = .012) and RSE and AER (β = 0.083; t = 3.02; p = .003) respectively.

The f-square effect size estimates were utilized to assess the meaningfulness of the path coefficients. Based on Cohen (1988)’s interpretation of f² values, all the path relations except the paths from RSE to AER and PRA, were not relevant in explaining variances in their respective constructs. R2 criteria was used to evaluate the predictive power of the structural model (Chin, 1998). The examination of the endogenous constructs’ predictive power shows that the primary outcomes exhibit moderate to very strong R² values. The results show that RSE, RAN, RU, and PRA explained 54.6% of the variance in AER. Similarly, RSE and RAN explained 36.6% of variation in PRA. Finally, RSE, RAN and PRA accounted for 37.8% variation in RU.

5. Discussion

In the absence of studies investigating the cognitive and emotional processes that influence students’ behavioral tendencies in research methods courses, this paper, within the theoretical perspectives of social cognitive theory and expectancy-value model, examined the interrelations among research anxiety, self-efficacy, positive research disposition, usefulness, and students’ academic effort in research methods.

Contrary to previous works (Macher et al., 2011; Maharajan et al., 2017) the association between research anxiety and academic effort in research was unsupported, this implies that anxiety levels of students did not influence the effort they expend in activities related to research methods. Perhaps, uneasiness about research does not seem to dissuade students from spending time on research methods. Again, results of the study did not provide support for the assumed relationship between anxiety and students’ positive disposition toward research methods thereby contradicting earlier works (Korobili et al., 2010; Akin & Kurbanglu, 2011). This implies that students’ apprehension about research method does not affect their positive disposition toward research methods. The hypothesized association between anxiety and perception of research usefulness was supported, thereby confirming previous studies (Callum et al., 2014; Liaw & Huang, 2013). This means that when students become anxious about research methods courses, they turn to consider research methods less important to their future career goals. Put differently, when students consider research as a useful course, their anxiety levels about research methods courses decreases. In confirmation of earlier studies (Munoz et al., 2018; Qudsyia & Putrib, 2016), relationship between research methods anxiety and self-efficacy was supported. This means that when students perceive themselves less capable in executing learning tasks related to research methods, their apprehension about the course increases. This finding can be explained by the idea of triadic reciprocal causation inherent in the SCT. Anxiety, which is a cognitive process within the person influencer in the triadic reciprocal determinism concept will relate to internal competences perceptions.

In tandem with existing literature (Iovu et al., 2015; Li, 2012), the more positive disposition students show towards research methods, the more effort they expend on research methods learning activities. Within the expectancy-value model, when students consider research as useful to their future plans, they are likely to develop positive disposition towards the course which leads to investment of effort into learning activities related to the course. Based on the results of the study, just as reported in previous studies (Guritno, & Siringoringo, 2013; Selevičienė & Burkšaitienė, 2015), perception of research usefulness increases with higher levels of positive disposition of students towards research. In other words, when students show favourable affect towards research, their perception of value of the course also increases. In support of the expectancy-value theory, results of the study indicated that as students’ perceived value of research increases, the more effort they expend on research methods learning activities, thereby supporting previous studies (Cole et al., 2008; Schmid & Bogner, 2015).

Based on the social cognitive theory, results of the study endorsed the proposition that self-beliefs about one’s competence in a subject drive academic effort. This finding demonstrates that students’ investment of time and resources into research learning activities enhances their perception of competencies in the subject. In other words, perception of self-efficacy leads to more effort and time spent in engaging with research methods learning activities. Relatedly, higher levels of students’ perceived self-efficacy in research methods enhance their positive disposition towards the course. In the perspective of expectancy-value model, results of the study indicate that students’ perception of competence in research heightens their value perception of research to the attainment of future career goals. Regarding the mediation analysis, the results show that students’ perception of research usefulness fully mediates the relationship between research anxiety and academic effort. The implication of this finding is that the extent to which anxiety reduces efforts in research depends on students’ perceived usefulness of research methods. Again, students’ perceived usefulness of research partially mediated the effect of positive attitude on the efforts students invest in research. With a negative coefficient value, the effect of attitude on effort increases or decreases depending on the value students place on research methods. Lastly, research usefulness partially mediated the association between self-efficacy and effort in research. This means that students’ perception of research usefulness enhances the effect of self-efficacy on students’ efforts in research.

5.1. Theoretical implications

Generally, undergraduate students’ anxiety studies have tended to focus on statistics, mathematics, test taking, and computer and ICT related issues, relative to research methods. Given that a substantial number of university students take a course in research methods prior to graduation, a study on research methods anxiety among undergraduate students is a worthwhile venture. This study therefore adds to the scanty knowledge on students’ research methods anxiety, self-efficacy, perception of research usefulness and academic effort. Furthermore, previous studies on undergraduates’ attitude toward research are largely atheoretical. However, this paper is among the few to have explicated a model of students’ research anxiety and attitude using SCT and EVT. Essentially, the results of the study indicate the applicability of both SCT and EVT to explain the effects of anxiety, attitude, self-efficacy and perception of usefulness on undergraduate students’ academic efforts in research methods. Furthermore, in view of the interrelations among the emotional, attitudinal and behavioral factors, the study results provide empirical support for the triadic reciprocal determinism in the SCT within the context of research teaching and learning in the domain of higher education milieu.

5.2. Practical implications

The results of study provide useful insights that have implications for both administrators and teaching faculty in higher education institutions. Considering the negative effect of anxiety on students’ perception of usefulness of research, teaching faculty should design and implement strategies that will help to reduce students’ anxiety levels towards research methods. Instructional strategies could be revised to reduce students’ anxiety levels. In this direction, adoption of student-centered approaches to facilitate teaching and learning of research methods is highly recommended. Lecturers should adopt inquiry-based and collaborating pedagogic strategies to teach research methods courses. These instructional strategies have the capacity to help reduce anxiety among students. To enable faculty, implement these instructional strategies, management of universities need to train faculty on these student-centered pedagogic approaches. Beyond instructional strategies, teaching faculty should improve relationships with students by making time for contact with students outside the classrooms. The use of tutorials for research courses should be encouraged as this forum will assist students outside the classroom thereby helping to reduce research methods anxiety among students. Furthermore, counselling units in universities should engage students to help reduce research anxiety. In this study, self-efficacy positively relates to students’ effort, positive disposition and perception of usefulness of research. It is therefore important to improve students’ self-efficacy that will have ripple effects on students’ academic effort in learning activities, develop positive attitudes toward the course as well as seeing research as important to their future careers and goals.

6. Limitations

There are limitations of the current study that need to be acknowledged or pointed out. First, the sample for the study was selected from a technical university in Ghana. By this single-case study, caution is advised when making generalization of the results of the study as selecting respondents from a single university limits the generalizability of the results of the study. Having said this, the results of the study are comparable to those conducted in other countries thereby reposing some level of confidence in the study. Secondly, respondents of the study were selected via accidental or convenience sampling technique. This approach raises questions about the representatives of the population from which the sample was selected.

7. Conclusions

Based on the perspectives of the social cognitive (SCT) and expectancy-value theories, this study investigated the interplay among research anxiety, positive disposition toward research, perception of usefulness, and self-efficacy, and academic effort in research. Research anxiety negatively affects perception of research usefulness and reduces students’ self-beliefs. Positive disposition towards research enhances efforts students expend on learning activities related to the course. Furthermore, self-efficacy drives students’ efforts, enhances positive disposition toward research and perception of research usefulness among students. Students’ perception of research usefulness mediates the relationship between anxiety and effort in research, positive research disposition and effort as well as research self-efficacy and effort in studies.

Data availability

All data generated or analyzed in relation to this study are available upon request.

Declarations

• Ethical approval: All procedures conducted in executing this study were in lined with institutional, national and international ethical standards involving human participants.

• Informed consent: The consent of all research participants was solicited before their participation in the study.

• Conflict of interest: The authors declare that they have no conflict of interest.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The authors received no direct funding for this research.