Megan C.
Connor
*a and
Jeffrey R.
Raker
*b
aDepartment of Chemistry and Biochemistry, Samford University, Birmingham, Alabama 35229, USA. E-mail: mconnor@samford.edu
bDepartment of Chemistry, University of South Florida, Tampa, Florida 33620, USA. E-mail: jraker@usf.edu
First published on 7th February 2024
Despite institutional reform efforts to increase use of evidence-based instructional practices (EBIPs) in undergraduate chemistry and STEM courses, didactic lecture remains the predominant mode of instruction. Research to inform these initiatives routinely focuses on drivers and barriers to EBIP adoption, with recent work investigating factors associated with faculty members’ cooperative adoption of EBIPs from five STEM disciplines including chemistry. To understand the role of these specific factors within undergraduate chemistry education across a broad set of institutions, we conducted a national survey of chemistry faculty members (n = 1105) from the United States in Spring 2023. The survey targeted constructs that may underlie the cooperative adoption of EBIPs, including faculty members’ perception of (1) using EBIPs as mutually beneficial, (2) having their success and failure intertwined, and (3) institutional climate around teaching. The survey also included items targeting teaching-specific social interactions, another potential aspect of cooperative adoption. Results from multilevel modeling suggest that EBIP adoption is associated with chemistry faculty members’ perception of using EBIPs as mutually beneficial, aligning with prior findings on STEM faculty members’ cooperative adoption of these practices. However, there is no evidence of an association between EBIP adoption and chemistry faculty members’ perception of campus climate around teaching, where prior findings indicate an inverse association among STEM faculty members. Results further indicate that EBIP adoption is associated with the number of people with whom one specifically discusses pedagogy, instruction, and assessment. Collectively, our results demonstrate that differences exist between STEM disciplines and point toward the chemistry education research community's responsibility to further explore EBIP adoption from a disciplinary lens. Our investigation also provides insight into factors associated with the cooperative adoption of EBIPs among chemistry faculty members on a national level; we identify several implications for how chemistry faculty member change agents (e.g., course coordinators, department leaders) may effectively promote EBIP adoption across the undergraduate chemistry curriculum.
Research further suggests that, across various STEM disciplines (e.g., chemistry, biology, physics, geology, environmental engineering, etc.), faculty members who are formal mentors play an essential role in the spread of knowledge surrounding EBIPs, serving as catalysts that help communities of practice dedicated to EBIP adoption grow and connect with one another (Ma et al., 2018; Mestre et al., 2019). Further, chemistry, physics, mathematics, and engineering faculty members who have more teaching-specific interactions tend to use student-centered (i.e., peer-led team learning, PLTL) instructional approaches, in contrast to those with fewer interactions who tend to use instructor-centered approaches (e.g., lecturing; Middleton et al., 2022). These faculty members also report greater experience with assessing learning, an essential aspect of evidence-based instruction (McConnell et al., 2020). However, those who use EBIPs tend to preferentially interact with other faculty members who use EBIPs, suggesting knowledge of EBIPs may remain with existing users (Lane et al., 2020).
EBIP adoption can, thus, be a cooperative rather than independent endeavor (McAlpin et al., 2022). In a recent study, McAlpin et al. (2022) investigated potential factors associated with STEM faculty members’ cooperative adoption of EBIPs, including their teaching-specific interactions (McAlpin et al., 2022); their investigation focused on faculty members across five STEM disciplines (i.e., biological sciences, chemistry, earth sciences, mathematics, and physics) at three large public research universities. Findings from this study suggest that faculty members are more likely to adopt EBIPs when they perceive such practices as mutually beneficial to themselves and their colleagues (McAlpin et al., 2022). Faculty members who are more often the target of teaching discussions, or those who serve as opinion leaders, also tend to be EBIP adopters. Additional research demonstrates that faculty members may serve as opinion leaders on teaching regardless of their academic rank (Lane et al., 2019). Training faculty members of any rank or tenure status within a department to use EBIPs may, thus, be one approach to supporting EBIP adoption. Moreover, faculty members who more strongly perceive their institutional climate as supportive of teaching innovation are also those who tend to not adopt EBIPs (McAlpin et al., 2022); while the nature of this relationship is unclear, it is possible that a population of STEM faculty members views institutional readiness for teaching innovation as, instead, institutional pressure to use EBIPs (McAlpin et al., 2022). Results of the investigation imply a need for activities designed to convince faculty of the mutual benefit of adopting EBIPs, which are distinct from activities that focus on teaching faculty how to implement EBIPs (McAlpin et al., 2022). Authors of the investigation also call for additional studies focused on factors underlying the cooperative adoption, as they may vary with context (McAlpin et al., 2022).
Further, the CACAO model prioritizes understanding potential adopters, as this insight enables change agents to effectively promote and support change (Dormant, 2011). The model identifies several aspects of an ideal change initiative from the perspective of potential adopters. These include the relative advantage and social impact of adopting a change, as well as the compatibility, simplicity, and adaptability of the change. Changes that afford additional benefits relative to the status quo are conducive to widespread adoption, whereas changes that offer little relative advantage are less likely to be widely adopted. Likewise, changes that positively impact the social relations of adopters are conducive to widespread adoption, whereas changes that would negatively impact their social relations are less likely to be adopted. The compatibility of the change with an organization's current systems and practices, its simplicity to implement, and its adaptability to adopters’ unique context further supports broad adoption. Conversely, change that conflicts with current systems and practices, is complex or complicated to implement, or cannot be tailored to one's context is less likely to be widely adopted (Dormant, 2011). When considering aspects of an ideal change initiative in the context of chemistry instructional reform, adoption of EBIPs would be most widespread (1) when chemistry faculty members perceive such practices as advantageous compared to other forms of instruction, (2) when they perceive EBIP adoption as having a positive (or at least neutral) impact on relationships with others in their department or institution, and (3) when they perceive EBIPs as compatible with current instructional approaches at their institution, easy to implement, and adaptable to their own classroom, chemistry content, etc. (McAlpin et al., 2022).
Previous results obtained from the CAFI provide initial insight into factors associated with STEM faculty members’ cooperative adoption of EBIPs (McAlpin et al., 2022), providing implications for how change agents may effectively promote and support their use. Namely, results suggest that perceiving EBIP adoption as a strategic complement is associated with EBIP adoption, whereas perceiving institutional readiness for change is inversely associated with adoption (McAlpin et al., 2022). The former association suggests that when faculty members believe that more people using EBIPs results in greater benefits, they are more likely to adopt EBIPs. The latter association suggests that faculty members who do not use EBIPs may perceive institutional readiness for change (and possibly institutional pressure to use EBIPs) more acutely than faculty members who already use EBIPs in their courses, though the nature of this relationship is unclear (McAlpin et al., 2022). However, the degree to which these associations broadly persist among postsecondary chemistry faculty members across institutions is unresolved. By using the CAFI to investigate the association of these factors with chemistry faculty members’ EBIP adoption on a national scale, researchers can obtain insight into the extent to which associations observed by McAlpin et al. (2022) extend to this instructor population. Such results will help change agents across institutions (e.g., chemistry department leaders) develop effective strategies for promoting and supporting reform across the undergraduate chemistry curriculum.
This finding adds to the growing body of research on how social interactions may influence EBIP adoption. However, such relationships are complex, resulting in numerous aspects about their nature which are unresolved. These include but are not limited to the nature of teaching discussions (e.g., which aspects of teaching are discussed) and engagement in teaching evaluations with one's discussion partner. Results from one investigation across three science departments (biology, chemistry, and geoscience) suggest that both the context and content of teaching conversations may be important for promoting EBIP adoption (Lane et al., 2022). While data were not disaggregated by discipline, office location and course overlap were found to connect faculty and supported the sharing of innovative teaching knowledge. Conservations that participants considered as influencing their use of EBIPs then focused on a range of topics, including course delivery, content coverage, teaching strategies, and the degree of course synchronization, among others (Lane et al., 2022). Moreover, it is possible that, while all discussions surrounding teaching are useful, only discussions about particular aspects of teaching drive EBIP adoption. Such insight would help opinion leaders focus their discussions on topics that will effectively promote reform. It is further possible that EBIP adoption is associated with having a faculty discussion partner that has observed one's teaching, or vice versa, providing departments with a formal structure for promoting EBIP use. Research demonstrates that faculty mentors may catalyze instructional change (e.g., Mestre et al., 2019); insight into the association between having a discussion partner who has observed one's teaching and EBIP adoption would thus provide a nuanced perspective into how mentors effectively promote reform.
1. To what extent are chemistry faculty members’ perceptions of strategic complements, interdependence, and campus climate around teaching associated with EBIP adoption in their postsecondary courses?
2. To what extent are chemistry faculty members’ social interactions relating to teaching associated with EBIP adoption in their postsecondary courses?
Institutional control | Highest chem. degree awarded | Number of institutions | Number of institutions with ACS approval | Number of faculty members | Number of respondents | Response rate (%) |
---|---|---|---|---|---|---|
Public | Bachelor's | 250 | 165 | 2664 | 334 | 12.5 |
Public | Graduate | 223 | 217 | 5982 | 548 | 9.2 |
Private | Bachelor's | 575 | 233 | 3558 | 488 | 13.7 |
Private | Graduate | 89 | 78 | 1912 | 145 | 7.6 |
Totals | 1137 | 693 | 14116 | 1515 | 10.7 |
Institutions in the United States awarding graduate degrees (i.e., master's or doctorate) typically have higher research activity compared to institutions awarding only bachelor's degrees. These institutions tend to have larger course sizes and faculty members with time-consuming research obligations (Cox et al., 2011), both of which can act as barriers to EBIP adoption (Lund and Stains, 2015; Shadle et al., 2017). Further, ACS is a professional society in the United States, and the Committee on Professional Training within ACS establishes evaluation criteria for undergraduate chemistry degree programs (Committee on Professional Training, 2023). Degree programs meeting evaluation criteria are granted approval from the ACS to award certified bachelor's chemistry degrees. ACS is analogous to the Royal Society of Chemistry (RSC) in the United Kingdom and the Royal Australian Chemical Institute (RACI) in Australia, where the ACS approval process is similar to the RSC and RACI accreditation processes. Approval from ACS to award certified bachelor's degrees has been found to be associated with greater use of evidence-based pedagogies (e.g., Connor et al., 2022).
The survey included previously published versions of the CAFI (see Appendix 1; McAlpin et al., 2022) and EBIP Adoption Scale (see Appendix 1; Landrum et al., 2017). Items targeting faculty members’ social interactions surrounding aspects of teaching and learning were developed using recommended items for collecting network data in the social sciences (Burt, 1984). Mirroring these recommended items (Burt, 1984), participants were first asked to list up to five people with whom they discussed teaching and learning over the past year (see Appendix 1). Subsequent items then probed different aspects of their interactions with each listed person (e.g., whether the listed person has observed their teaching, or vice versa; whether the listed person engages in discipline-based education research; etc.; see Appendix 1). Literature on collecting network data recommends that survey respondents list and respond in reference to a maximum of five people; this maximum accounts for the average number of people that respondents typically list, time considerations for completing the survey, the increased likelihood of survey fatigue when responding in reference to more than five individuals, and evidence on the amount of information individuals can comfortably retain at once in their memory (Miller, 1956; Simon, 1974; Burt, 1984).
Further, social interaction items were designed to gain general rather than detailed insight into the nature of social interactions. For example, participants were simply asked how many people with whom they discussed teaching and learning over the past year, with the anticipation that some participants would restrict their responses to people with whom they had frequent or focused interactions while others would not. This decision was informed by the study's aim of investigating various factors underlying the cooperative adoption of EBIPs in addition to social interaction variables. Numerous items targeting specific aspects of respondents’ social interactions would contribute to survey fatigue given the already large number of probed variables; to reduce survey fatigue while still investigating a range of factors underlying cooperative adoption, items thus only targeted general aspects of social interactions. By identifying general social interaction variables associated with the cooperative adoption of EBIPs, this study aims to provide a foundation through which future studies can more deeply probe specific aspects. Relevant survey items are provided in Appendix 1 in the order in which they were presented to participants.
Institutional control | Highest chem. degree awarded | Number of institutions in sample | Number of institutions with ACS approval in sample | Number of faculty members in sample | Response rate (%) |
---|---|---|---|---|---|
Public | Bachelor's | 122 | 102 | 240 | 9.0 |
Public | Graduate | 170 | 168 | 382 | 6.4 |
Private | Bachelor's | 234 | 136 | 375 | 10.5 |
Private | Graduate | 61 | 55 | 108 | 5.6 |
Totals | 587 | 461 | 1105 | 7.8 |
EBIP adoption scale item | Score (number of “yes” responses) | CACAO adoption stage | Modified CACAO adoption stage |
---|---|---|---|
0 | Awareness | Awareness | |
Prior to this survey, I already knew about evidence-based instructional practices (EBIPs). | 1 | Awareness | Awareness |
I have thought about how to implement EBIPs in my courses. | 2 | Mental tryout | Tryout |
I have spent time learning about EBIPs (e.g., attended workshops, experimented in class, read education literature), and I am prepared to use EBIPs. | 3 | Hands-on tryout | Tryout |
I consistently use EBIPs in my courses. | 4 | Adoption | Adoption |
I consistently use EBIPs, and I continue to learn about and experiment with new EBIPs. | 5 | Adoption | Adoption |
I have evidence that my teaching has improved since I started using EBIPs. | 6 | Adoption | Adoption |
Various statistics are used to evaluate the reliability and unidimensionality of measures obtained via Guttman scales, including the coefficient of reproducibility (CR), coefficient of scalability (CS), minimal marginal reproducibility (MMR), and percent improvement (PI; McIver and Carmines, 1981). Ideally, CR values should be greater than 0.90 and CS values greater than 0.60 for evidence of reliability and unidimensionality (Abdi, 2010). Lower MMR values and higher PI values are further ideal. For measures obtained using the EBIP Adoption Scale for the study herein, CR = 0.96 and CS = 0.85. These values exceed the cutoff criteria, providing evidence of reliability and unidimensionality. Further, MMR = 0.74 and PI = 0.22. The MMR and PI are higher and lower than desired, respectively, though not prohibitive. With evidence of reliability and unidimensionality, participants’ scores on the EBIP adoption scale were then determined using their total number of “yes” responses (see Table 3). Items from the EBIP Adoption Scale and respondents’ total scores were further mapped onto modified CACAO adoption stages as done by Yik et al. (2022); (see Table 3) to facilitate multilevel modeling. In these modified stages, Mental Tryout and Hands-on Tryout are condensed into one stage, i.e., Tryout. Participants were thus categorized into one of three EBIP adoption stages: Awareness, Tryout, or Adoption.
To address Research Question 2, two additional multilevel binary logistic regression models (i.e., Tryout – Social and Adoption – Social) are used to investigate the extent to which faculty members’ social interactions relating to teaching and learning are associated with EBIP adoption when accounting for perceptions of strategic complements, interdependence, and climate. The Tryout – CAFI model is used to distinguish between awareness and tryout, and the Adoption – CAFI model is used to distinguish between tryout and adoption. Two multilevel binary logistic regression models are again used to model two different outcomes (i.e., tryout and adoption) to obtain the most parsimonious and interpretable regression coefficients.
A multilevel approach is used to account for the nesting of faculty members within departments, a violation of the assumption that data represent independent observations (see Raudenbush and Bryk, 2002). All models included two levels, with faculty members at the first level and departments at the second level. The number of departments corresponds to the number of institutions in the study sample (see Table 2), as each institution has a single chemistry department. The Tryout (CAFI and Social) models included n = 322 participants and n = 290 groups (i.e., departments). The Adoption (CAFI and Social) models included n = 997 participants and n = 550 groups. Mean and variance adaptive Gauss–Hermite quadrature (mvaghermite) integration with seven integration points was used for modeling with the “melogit” command in StataSE Version 17 (StataCorp, 2021).
Adoption stage | Number of study participants (n = 1105) | Percentage of study participants (%) |
---|---|---|
Awareness | 108 | 9.8 |
Tryout | 258 | 23.3 |
Adoption | 739 | 66.9 |
Within this distribution, 66.9% of study participants are in the Adoption stage (see Table 4). This percentage is substantive and potentially an anomaly, as research suggests lecture-based approaches remain the dominant mode of instruction in undergraduate STEM courses (Stains et al., 2018). Given that only faculty members who provided complete responses to factors underlying the cooperative adoption of EBIPs were included in the study sample, it is possible that study participants are committed to teaching and are, in turn, more likely to be in the Adoption stage. However, a distribution of stages needed to investigate factors associated with cooperative adoption was obtained (see Table 4), so this potential response bias does not prohibit the investigation.
Predictor variable | Definition | Coding |
---|---|---|
a DBER investigates learning and teaching in a discipline from a perspective reflecting disciplinary knowledge and practices, and it is informed by research on learning and cognition (National Research Council, 2012). SoTL investigates learning and teaching with an emphasis on developing reflective practice and using classroom-based evidence (National Research Council, 2012), though it does not emerge from education theory or the learning sciences (Coppola and Krajcik, 2013). | ||
Department demographic variables | ||
ACS approval | Approval from ACS to award certified bachelor's chemistry degrees | 0 = no, 1 = yes |
Highest chemistry degree awarded – bachelor's | A chemistry bachelor's degree is the highest degree awarded by a department | 0 = no, 1 = yes |
CAFI variables | ||
Strategic complements | CAFI factor scores from CFA | |
Interdependence | CAFI factor scores from CFA | |
Climate | CAFI factor scores from CFA | |
Social interaction variables | ||
Total people | Total number of people with whom the faculty member discussed teaching and learning over the past year | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Department | Of the people with whom the faculty member discussed teaching and learning over the past year, the number that were from their department | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
College | Of the people with whom the faculty member discussed teaching and learning over the past year, the number that were from their college or university | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
External | Of the people with whom the faculty member discussed teaching and learning over the past year, the number the faculty member has observed teach during that time | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Observed by | Of the people with whom the faculty member discussed teaching and learning over the past year, the number that observed the faculty member's teaching during that time | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Observed | Of the people with whom the faculty member discussed teaching and learning over the past year, the number the faculty member observed teaching during that time | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
SoTL or DBER | Of the people with whom the faculty member discussed teaching and learning over the past year, the number engaged in Scholarship of Teaching and Learning (SoTL) or Discipline-Based Education Research (DBER)a | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Textbook | Of the people with whom the faculty member discussed teaching and learning over the past year, the number they discussed the textbook with | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Content | Of the people with whom the faculty member discussed teaching and learning over the past year, the number they discussed course content with | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Pedagogy and instruction | Of the people with whom the faculty member discussed teaching and learning over the past year, the number they discussed pedagogy and instruction with | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Assessment | Of the people with whom the faculty member discussed teaching and learning over the past year, the number they discussed assessment with | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
Academic dishonesty and integrity | Of the people with whom the faculty member discussed teaching and learning over the past year, the number they discussed academic dishonesty and integrity with | 0 = no people, 1 = 1 person, 2 = 2 people, 3 = 3 people, 4 = 4 people, 5 = 5 people |
RQ1: To what extent are chemistry faculty members’ perceptions of strategic complements, interdependence, and campus climate around teaching associated with EBIP adoption in their postsecondary courses?
ORs for the Tryout – CAFI model range from 1.02 to 2.78, and those for the Adoption – CAFI model range from 0.88 to 2.82 (see Table 6). Strategic complements is a significant predictor variable in both models. The OR = 1.53 for this variable in the Tryout – CAFI model, meaning the odds of a chemistry faculty member being in the Tryout versus Awareness stage are 1.53 times greater for one standard deviation above the mean when the faculty member perceives EBIP adoption as a strategic complement. The OR = 2.82 in the Adoption – CAFI model, indicating the odds of a faculty member being in the Adoption versus Tryout stage are 2.82 times greater for one standard deviation above the mean when the faculty member perceives EBIP adoption as such a complement. Other CAFI variables (i.e., interdependence and climate) are nonsignificant predictors with ORs approaching one across models; there is, thus, no evidence of an association between these variables and EBIP adoption.
Variable | EBIPs tryout | EBIPs adoption | ||||
---|---|---|---|---|---|---|
OR | SE | p | OR | SE | p | |
*Significance at 0.05 level, ** significance at 0.01 level, and *** significance at <0.001 level. | ||||||
Departmental demographic variables | ||||||
ACS approval | 1.62 | 0.58 | 0.182 | 1.06 | 0.27 | 0.822 |
Highest degree awarded – bachelor's | 1.78 | 0.45 | 0.025* | 1.19 | 0.22 | 0.351 |
CAFI variables | ||||||
Strategic complements | 1.53 | 0.18 | <0.001*** | 2.82 | 0.30 | <0.001*** |
Interdependence | 1.26 | 0.19 | 0.136 | 0.88 | 0.10 | 0.275 |
Climate | 1.02 | 0.12 | 0.858 | 1.03 | 0.09 | 0.651 |
RQ2: To what extent are chemistry faculty members’ social interactions relating to teaching associated with EBIP adoption in their postsecondary courses?
For the Tryout – Social model, ORs ranged from 0.84 to 1.75 (see Table 7). Strategic complements is again a predictor variable in this more comprehensive model, with an OR = 1.61 approaching that of the Tryout – CAFI model. No social interaction variables were significant predictors, with p-values greater than or equal to 0.115 and ORs near one. There is, thus, no evidence that faculty members’ social interactions relating to teaching and learning are associated with being in the Tryout versus Awareness stage, or vice versa.
Variable | EBIPs tryout | EBIPs adoption | ||||
---|---|---|---|---|---|---|
OR | SE | p | OR | SE | p | |
*Significance at 0.05 level, ** significance at 0.01 level, and *** significance at <0.001 level. | ||||||
Departmental demographic variables | ||||||
ACS approval | 1.75 | 0.71 | 0.170 | 1.05 | 0.28 | 0.839 |
Highest degree awarded – bachelor's | 1.47 | 0.43 | 0.185 | 1.06 | 0.21 | 0.776 |
CAFI variables | ||||||
Strategic complements | 1.61 | 0.24 | 0.002** | 2.80 | 0.30 | <0.001*** |
Interdependence | 1.10 | 0.20 | 0.579 | 0.83 | 0.10 | 0.125 |
Climate | 1.00 | 0.13 | 0.971 | 1.02 | 0.09 | 0.809 |
Social interaction variables | ||||||
Total people | 1.23 | 0.24 | 0.284 | 1.06 | 0.12 | 0.628 |
Department | 0.95 | 0.21 | 0.805 | 0.79 | 0.11 | 0.088 |
College | 1.29 | 0.39 | 0.396 | 0.95 | 0.16 | 0.782 |
External | 0.86 | 0.30 | 0.664 | 0.84 | 0.17 | 0.409 |
Observed by | 0.97 | 0.20 | 0.867 | 1.16 | 0.14 | 0.206 |
Observed | 1.01 | 0.17 | 0.961 | 0.86 | 0.09 | 0.160 |
SOTL or DBER | 0.94 | 0.10 | 0.576 | 1.04 | 0.07 | 0.617 |
Textbook | 1.04 | 0.13 | 0.779 | 0.88 | 0.07 | 0.133 |
Content | 0.97 | 0.14 | 0.804 | 1.16 | 0.10 | 0.077 |
Pedagogy and instruction | 1.21 | 0.15 | 0.124 | 1.23 | 0.10 | 0.014* |
Assessment | 0.84 | 0.12 | 0.239 | 1.18 | 0.10 | 0.044* |
Academic dishonesty and integrity | 1.21 | 0.15 | 0.115 | 0.94 | 0.06 | 0.353 |
ORs for the Adoption – Social model range from 0.79 to 2.80 (see Table 7). The largest OR (i.e., 2.80) corresponds to strategic complements, approaching the OR for this variable in the Tryout – CAFI model. Two social interaction variables are significant predictors of being in the Adoption versus Tryout stage: the total number of people with whom a chemistry faculty member discusses pedagogy and instruction (OR = 1.23) and the total number of people with whom they discuss assessment (OR = 1.18). The odds of a chemistry faculty member being in the Adoption versus Tryout stage, thus, increases by 1.23 and 1.18 times, respectively, for each additional person with whom they discuss pedagogy/instruction and assessment. Notably, the total number of people with whom a faculty member discusses teaching and learning in general is not a significant predictor of being in the Adoption versus Tryout stage. Other social interaction variables were nonsignificant predictors in this model, with ORs approaching one.
To further understand EBIP adoption across the undergraduate chemistry curriculum, additional research is needed to identify specific mutual benefits that act as drivers of cooperative adoption; for example, what mutual benefits does cohort-based professional development involving multiple members of the same chemistry department afford, and what role do these benefits play in supporting EBIP adoption? Chemistry faculty members in the United States also report pressure from various chemistry communities (e.g., textbook authors, ACS exam authors, and department colleagues) to cover a breadth of chemistry content rather than adopting a depth approach better aligned with how people learn (Kraft et al., 2023). These content coverage expectations also function as a barrier to EBIP adoption (Shadle et al., 2017). Thus, research could also investigate whether more chemistry department colleagues using EBIPs functions to reduce this pressure (i.e., a mutual benefit) and, in turn, further drive EBIP use.
The association between EBIP adoption and chemistry faculty members’ perception of using EBIPs as mutually beneficial aligns with prior findings on STEM faculty members’ cooperative adoption of EBIPs (McAlpin et al., 2022). However, there is no evidence of an association between chemistry faculty members’ perception of campus climate around teaching and EBIP adoption, where McAlpin et al. (2022) observed an inverse association between STEM faculty members’ perception of campus climate and adoption. Conversely, prior research demonstrates that chemistry faculty members’ perception of departmental climate around teaching is associated with EBIP adoption (Connor and Raker, 2023), while no evidence of an association is found for STEM faculty members’ perception of departmental climate and adoption (Emery et al., 2021; Shi and Stains, 2021). These differences suggest further exploration is needed to understand the role of departmental versus campus climate in the adoption of EBIPs among chemistry faculty members, including the possibility that their instructional practices are distinctly influenced by their perception of departmental rather than campus climate when compared to faculty members in other STEM disciplines.
Further, we found no evidence of an association between the total number of people with whom a faculty member discusses teaching regardless of topic and EBIP adoption. Collectively, these results suggest that while talking to others about teaching is important, it is the topics of the conversations that are associated with sustained adoption among chemistry faculty members. Additional studies focused on the nature of these topic-specific discussions, including frequencies, contexts, and objectives, will thus be essential for further promoting EBIP adoption in postsecondary chemistry courses. Future research on EBIP adoption in postsecondary chemistry courses should also routinely consider social interactions, as these findings suggest that focusing only on individual chemistry faculty members and their enacted teaching practices is insufficient for exploring how and why EBIPs are adopted.
Results also suggest that, in addition to the important mindset associated with trying out EBIPs, chemistry faculty members need communities in which these topic-specific conversations can occur for sustained EBIP adoption. There is no evidence of an association between EBIP adoption and the number of people in one's department, college, or university with whom they discuss teaching and learning. This result differs from that of another study focused on faculty members from multiple STEM disciplines (including chemistry, though data were not disaggregated by discipline), which found that EBIP adoption is associated with having both deeper and more extensive social connections within and across departments (Middleton et al., 2022). For chemistry faculty members, the composition of these communities may, therefore, be less important than the topics of their conversations. Chemistry faculty learning communities (Houseknecht et al., 2020), communities of practice (CoPs) (Raker et al., 2020), and departmental curriculum committees could, thus, provide the space to initiate and catalyze such discussions. A number of chemistry-specific CoPs have already formed to provide such spaces and, in turn, support EBIP adoption across institutions. For instance, OrganicERs is a CoP designed to introduce active-learning techniques to organic chemistry faculty members across the United States (Leontyev et al., 2020). Stone et al. (2020) also report the formation of a CoP to support chemistry faculty members in implementing course-based undergraduate research experiences, an evidence-based pedagogy for the instructional laboratory, across institutions. Further, the Interactive Online Network of Inorganic Chemists (IONiC) is a CoP committed to sharing instructional content and evidence-based teaching practices among inorganic chemistry faculty members internationally (Watson et al., 2020). Moreover, organically emerging discussions about these topics may be just as important, though additional research is needed to understand specific aspects of these conversations.
When conceptualizing the work of EBIP developers, evaluators, and disseminators, it is important to distinguish between building awareness, trying out EBIPs, and ultimately adopting EBIPs (Landrum et al., 2017). The work herein corroborates work from others that that there may not be a single approach to catalyze awareness, tryout, and adoption, as these groups are not the same (Viskupic et al., 2022; Yik et al., 2022a). Specifically, results suggest that chemistry faculty members’ mindset about the mutual benefits of using EBIPs may be an initial, important step on the path of adoption. Communities in which chemistry faculty members can discuss instruction/pedagogy and assessment may then be essential for moving individuals beyond the tryout stage.
Further, a number of our findings are distinct from those of studies focused on faculty members from other STEM disciplines or multiple (though not disaggregated) STEM disciplines. These differences suggest that effective approaches to catalyzing awareness, tryout, and adoption are unique to not only adoption stage but also to discipline. These key differences are highlighted and reemphasized below:
• We found no evidence of an association between chemistry faculty members’ perception of campus climate around teaching and EBIP adoption, though McAlpin et al. (2022) observed an inverse association between STEM faculty members’ perception of campus climate and adoption. When combined with prior research demonstrating that chemistry faculty members’ perception of departmental climate around teaching is associated with EBIP adoption (Connor and Raker, 2023), this finding suggests that chemistry faculty members’ instructional practices are distinctly influenced by their perception of departmental rather than campus climate when compared to faculty members in other STEM disciplines. Influencing chemistry faculty members’ perceptions of departmental rather than institutional climate may, thus, be a more productive approach to increase EBIP use.
• We found no evidence of an association between the number of people with whom a chemistry faculty member discusses course content and EBIP adoption, though in an investigation by Lane et al. (2022), faculty members from multiple STEM disciplines report that their conversations about course content influence their use of EBIPs. Content coverage expectations are a barrier to EBIP adoption (Shadle et al., 2017), and chemistry faculty members report pressure from multiple chemistry communities to cover a breadth of content (Kraft et al., 2023). This finding raises the possibility that chemistry faculty members choose to avoid this topic in conversations with colleagues or that conversations surrounding content coverage are not productive. This issue merits further research, as absent or futile conversations surrounding content coverage may be impeding EBIP adoption in undergraduate chemistry education.
• We found no evidence of an association between chemistry faculty members’ EBIP adoption and the number of people in their department, college, or university with whom they discuss teaching and learning, though in another study targeting faculty members from multiple STEM disciplines (Middleton et al., 2022), EBIP adoption is associated with having both deeper and more extensive social connections within and across departments. This difference suggests that while chemistry faculty members need communities in which topic-specific conversations about teaching can occur for sustained EBIP adoption, unlike in other STEM disciplines, the composition of these communities may not be essential. Chemistry faculty members may, therefore, uniquely benefit from participating in existing chemistry-specific faculty learning communities and CoPs that aim to support EBIP use across institutions. It is further possible that chemistry faculty members have yet to develop extensive, teaching-focused social networks within their departments and institutions when compared to faculty members from other STEM disciplines, potentially due to insufficient reform efforts targeting these individuals. Future research should address underlying causes for this possible lack of association, as expanding these social networks among chemistry faculty members could effectively promote EBIP adoption in undergraduate chemistry education.
First, the social dimensions explored could be strengthened by a social network analysis approach, as has been used in other studies of faculty members’ teaching-specific interactions (Kezar, 2016; Lane et al., 2020; McConnell et al., 2020). In a context where a complete social network can be created, a more robust analysis of social influences on outcome measures can be conducted. For example, McAlpin et al. (2022) used indegree from a social network analysis as a measure of opinion leadership when exploring the relationship between cooperative adoption factors and EBIP adoption. However, to conduct such a study, one is limited to smaller sample sizes and fewer departmental/institution sites; McAlpin et al. (2022) only studied three institutions and five STEM disciplines within each institution. This limits generalizability and estimates of the relationship of interest at a national level, as was the goal of the study herein. There is a possible ego network approach for doing such work (Arnaboldi et al., 2012); however, given the broad focus of the study herein, there are a limited the number of questions that could be expected to be answered by a respondent in a reasonable time period. Results of the study herein suggest that there are points of interest that should be considered in designing and executing a rigorous ego network analysis investigation.
Second, the work herein overlaps with multiple survey research studies in chemistry and STEM for which a growing number of important factors of association with EBIP adoption have been identified (Gibbons et al., 2017; Raker et al., 2021a; Yik et al., 2022a, 2022b). It is unmanageable to conduct a study with sufficient statistical power to evaluate all possible associations with EBIP adoption. Still, if the focus of such work is to identify levers for change, it is prudent to suggest that future studies incorporate previously found factors into such studies. For example, a number of studies show that classroom setup (i.e., classroom space conducive to small group work) is associated with decreased time lecturing and sustained EBIP adoption (Yik et al., 2022a, 2022b). Thus, it is important to use this potentially confounding factor in studies evaluating the effectiveness of a professional development intervention on adoption of EBIPs.
Third, the distribution of study participants’ EBIP adoption stage may not reflect that of the larger study population, as research suggests lecture-based approaches remain the dominant mode of instruction in undergraduate STEM courses (Stains et al., 2018). In a representative sample, the majority of participants would, thus, likely/hypothetically be in the Awareness or Tryout stages. The relatively small percentage of study participants in the Awareness stage may mean that our data are not representative of all chemistry faculty members in this adoption stage. The smaller number of participants in this stage may have also limited the statistical power and, in turn, ability to identify factors associated with the tryout of EBIPs using the Tryout – CAFI and Tryout – Social logistic regression models. Therefore, additional factors may be associated with the tryout of EBIPs, though this investigation is unable to provide evidence of these associations. However, the large number of study participants in the Adoption stage, combined with the moderate number in the Tryout stage, allowed for the identification of multiple factors associated with adopting EBIPs.
Yes | No | |
1. Prior to this survey, I already knew about evidence-based instructional practices (EBIPs) | ○ | ○ |
2. I have thought about how to implement EBIPs in my courses. | ○ | ○ |
3. I've spent time learning about EBIPs (e.g. attended workshop, experimented in class, read education literature) and I am prepared to use them. | ○ | ○ |
4. I consistently use EBIPs in my courses. | ○ | ○ |
5. I consistently use EBIPs and I continue to learn about and experiment with new EBIPs. | ○ | ○ |
6. I have evidence that my teaching has improved since I started using EBIPs. | ○ | ○ |
1. My department is more effective when all faculty are committed to using EBIPs.
Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
2. My students are likely to be more receptive to an EBIP in my course, if they have had a similar EBIP in another departmental course.
Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
3. I do not use EBIPs because I want students to be exposed to lecture-based teaching.
Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
4. Departments best foster student learning if EBIPs are adopted by all faculty members.
Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
5. It's easier for me to adopt EBIPs if someone else in my department is adopting them because my colleagues serve as a handy resource for me.
Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
6. I am convinced that EBIPs are of little value for student learning.
Strongly disagree | Disagree | Somewhat disagree | Neither agree nor disagree | Somewhat agree | Agree | Strongly agree |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
We are now going to ask you 6 questions about how you perceive yourself in relation to your closest departmental colleague.
1. When something good happens to a close colleague in my department, that is
Very bad for me | Bad for me | Somewhat bad for me | Neither good nor bad for me | Somewhat good for me | Good for me | Very good for me |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
2. When something bad happens to a close colleague in my department, that is:
Very bad for me | Bad for me | Somewhat bad for me | Neither good nor bad for me | Somewhat good for me | Good for me | Very good for me |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
3. When a close colleague in my department succeeds, I feel:
Very bad | Bad | Somewhat bad | Neither good nor bad | Somewhat good | Good | Very good |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
4. When a close colleague in my department fails, I feel:
Very bad | Bad | Somewhat bad | Neither good nor bad | Somewhat good | Good | Very good |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
5. My close departmental colleague's gain is:
My loss | Neither my gain or my loss | My gain | ||||
○ | ○ | ○ | ○ | ○ | ○ | ○ |
6. My close departmental colleague's loss is:
My loss | Neither my gain or my loss | My gain | ||||
○ | ○ | ○ | ○ | ○ | ○ | ○ |
For each item, please select the scale point that best represents your opinion.
Each statement begins with “I believe that… ”
the campus culture is generally supportive of teaching. | ○ | ○ | ○ | ○ | ○ | ○ | ○ | the campus culture is generally unsupportive of teaching. |
the campus culture is shaped by leaders who are not supportive of my teaching. | ○ | ○ | ○ | ○ | ○ | ○ | ○ | the campus culture is shaped by leaders who are supportive of my teaching. |
the campus culture breeds divisiveness in teaching discussions. | ○ | ○ | ○ | ○ | ○ | ○ | ○ | the campus culture breeds collaboration in teaching discussions. |
the campus culture does not value teaching. | ○ | ○ | ○ | ○ | ○ | ○ | ○ | the campus culture values teaching. |
the campus culture connects me with other teachers. | ○ | ○ | ○ | ○ | ○ | ○ | ○ | the campus culture isolates me from other teachers. |
1. How many people have you discussed teaching and learning with over the last year?
0 | 1 | 2 | 3 | 4 | 5 | Prefer not to disclose |
○ | ○ | ○ | ○ | ○ | ○ | ○ |
2. Please provide a number (e.g., “101”), letter (e.g., “AG”), or pseudonym (e.g., “Susan”; please do not provide their actual name) for each of your people.
3. For each of your people, what is their affiliation?
In your department | In your college or university | External to your college or university | |
Identifier #1 | ○ | ○ | ○ |
Identifier #2 | ○ | ○ | ○ |
… | ○ | ○ | ○ |
4. Of your people, who has observed your teaching in the last year? (select all that apply)
5. Of your people, who have you observed their teaching in the last year? (select all that apply)
6. To the best of your knowledge, who of your people have or are currently engaged in STEM education research or scholarship of teaching and learning? (select all that apply)
7. For each of your people, what topics related to teaching and learning have you discussed in the last year? (select all that apply)
Textbook | Course content | Pedagogy and instruction (e.g., group work, classroom response systems) | Assessment (e.g., examinations) | Academic dishonesty and cheating | |
Identifier #1 | ○ | ○ | ○ | ○ | ○ |
Identifier #2 | ○ | ○ | ○ | ○ | ○ |
… | ○ | ○ | ○ | ○ | ○ |
Demographic variables | Number of faculty members in sample (n = 1105) | Percentage of faculty members in sample (%) |
---|---|---|
Gender | ||
Man | 517 | 46.8 |
Woman | 425 | 38.5 |
Nonbinary | 12 | 1.1 |
Transgender | 7 | 0.6 |
Prefer not to disclose | 54 | 4.9 |
No response | 90 | 8.1 |
Race | ||
American Indian or Alaska Native only | 0 | 0.0 |
Asian only | 56 | 5.1 |
Black or African American only | 31 | 2.8 |
Native Hawaiian or Other Pacific Islander only | 0 | 0.0 |
White only | 822 | 74.4 |
Some other race, ethnicity, or origin only | 7 | 0.6 |
More than one race | 29 | 2.6 |
Prefer to self-describe | 15 | 1.4 |
Prefer not to disclose | 55 | 5.0 |
No response | 90 | 8.1 |
Factor | Items | Mean | SD | Skewness | Kurtosis | Loading |
---|---|---|---|---|---|---|
a Item was reverse coded prior to analysis, *p < 0.01, **p < 0.001. | ||||||
Complements | comp1 | 4.52 | 1.50 | −0.34 | 2.73 | 0.81 |
comp2 | 5.09 | 1.24 | −0.73 | 3.76 | 0.54 | |
comp3a | 5.67 | 1.45 | −1.05 | 3.36 | 0.53 | |
comp4 | 4.45 | 1.52 | −0.39 | 2.60 | 0.74 | |
comp5 | 4.71 | 1.56 | −0.64 | 2.81 | 0.41 | |
comp6a | 5.69 | 1.40 | −1.05 | 3.57 | 0.65 | |
Interdependence | inter1 | 5.37 | 1.05 | −0.25 | 2.64 | 0.82 |
inter2a | 5.20 | 0.99 | −0.07 | 2.43 | 0.50 | |
inter3 | 6.08 | 0.85 | −1.01 | 4.69 | 0.66 | |
inter4a | 5.86 | 0.83 | −0.69 | 4.26 | 0.47 | |
inter5 | 5.48 | 1.14 | −0.16 | 2.16 | 0.73 | |
inter6a | 5.44 | 1.11 | −0.15 | 2.32 | 0.57 | |
Climate | clim2a | 5.09 | 1.75 | −0.76 | 2.49 | 0.67 |
clim3 | 4.62 | 1.72 | −0.47 | 2.27 | 0.74 | |
clim4 | 4.74 | 1.58 | −0.57 | 2.67 | 0.60 | |
clim5 | 5.23 | 1.74 | −0.89 | 2.79 | 0.75 | |
clim6a | 4.49 | 1.65 | −0.33 | 2.22 | 0.52 | |
Covariance coefficient | ||||||
Complements and Interdependence | 0.17** | |||||
Complements and Climate | 0.12* | |||||
Interdependence and Climate | 0.21** |
Factor | α | ω |
---|---|---|
Complements | 0.79 | 0.89 |
Interdependence | 0.85 | 0.93 |
Climate | 0.79 | 0.83 |
Models | Number of participants | Number of groups (i.e., departments) | Minimum number of participants per group | Maximum number of participants per group | ICC |
---|---|---|---|---|---|
Tryout (CAFI and Social) | 366 | 290 | 1 | 9 | <0.01 |
Adoption (CAFI and Social) | 997 | 550 | 1 | 12 | 0.05 |
This journal is © The Royal Society of Chemistry 2024 |