Power’s influence on boundary-crossings between counselor educators and counselor education students

Abstract Counselor educators (CEs) hold legitimate power over counselor education students (CESs). It would be helpful for CEs to understand how power differentials influence boundary-crossing perceptions within the legitimate hierarchy of counselor education. CEs and CESs were separately shown the same 16 boundary-crossing scenarios that could occur between the groups. When CEs are negotiating boundary-crossings with CESs, Burian and Slimp’s model would likely be most helpful as it is specific to counselor education and requires CEs to contemplate motives when evaluating boundary-crossings. However, when CESs are negotiating boundary-crossing concerns with CEs, Gottlieb’s model would likely be most helpful. CESs in this study were very concerned about boundary-crossings that encroached upon their personal needs that were not a requirement for successfully graduating from a counselor education program. Consequently, CEs may want to become clearer in their understanding of the differences between optional behaviors versus those of mandated requirements for graduation. A CE and a CES sharing a car ride to a professional meeting is an optional boundary-crossing because it does not assist a CE in completing the mandated requirements for graduation. This contrasts with the mandated graduation requirement that a CE provides the required supervision hours to a CES.


PUBLIC INTEREST STATEMENT
The educational process results in professors holding power over graduate students. At the graduate level, professors often take on more than one role with a student serving as advisor, classroom instructor, and supervisor of the student's entry into the profession. The more roles that a professor takes with a student, the more likely that boundary-crossings will occur. Most often, boundary-crossings are harmless and productive. However, even with the best of intentions, they can result in the student experiencing confusion and stress. Researchers compared the rankings of 16 boundary-crossings that can occur between counselor education students (CESs) and counselor educators (CEs). The results suggested that CESs were not worried about boundarycrossings when CEs were not meeting their professional needs. However, they were unhappy when their personal needs were negatively impacted. CEs were concerned when CESs' needs were negatively affected by roles that are not a requirement for graduation.

Power in legitimate hierarchies
Power is the ability to asymmetrically influence others and distribute resources as part of a professor's position of authority and possession of expertise (Bolkan & Goodboy, 2013;Goodboy et al., 2011). A professor's power is customarily obtained when students without power follow the directives of a professor (Graham, 2015;Heator, 2018). These types of power differentials simplify and organize social exchanges by quickly helping people determine appropriate responses to others (Yap et al., 2013). Students must be willing to give their power away to professors, both groups must be willing to justify the disparity in power, and both groups must be willing to be mutually dependent upon the other group (Bolkan & Goodboy, 2013;Goodboy et al., 2011;Lammers et al., 2010). As such, the relationship between students and professors as well as professors' control over resources serves as the basis for the exchange of power between CEs and CESs (Heator, 2018). Hays and Goldstein (2015) stated that power is more easily accepted and viewed as appropriate and deserved when there is a legitimate hierarchy between the groups. A legitimate position of power comes with the right to openly express personal opinions and values, instruct the group without power on acceptable and unacceptable behavior, and be angry with the group without power when they do not follow directions (Lammers et al., 2010). Once an individual has obtained power through their legitimate position of authority or possession of expertise, the individual does not need the support of others to maintain that power (Hays & Goldstein, 2015).

The legitimately powerful
As the powerful can adopt a stance of entitlement, hypocritical behavior has the potential to flourish. Lammers et al. (2010) stated that hypocritical behavior occurs when a powerful individual espouses upholding strict moral standards in themselves and others, but that powerful individual violates those same moral standards in their behavior. They found that moral hypocrisy only occurs when the powerful hold legitimate power over the powerless. Lammers et al. (2010) also found that groups with power in legitimate hierarchies most often hold stricter moral standards of the powerless than of themselves, which can allow the powerful to easily justify and excuse immoral behaviors towards the powerless (Lammers et al., 2010).
Consequently, there are inherent dangers of exploitation for the powerless in legitimate hierarchies. This occurs because the powerful have the potential to concentrate on self-interest by focusing on the possible reward of any action (Sharabi-Levine, 2016;Smith & Bargh, 2008). This can lead the powerful to be presumptuous and more confident in their behaviors, which ultimately can result in distancing themselves from the perspective of the powerless (Lammers et al., 2008;Magee & Smith, 2013). As such, feelings of entitlement, a reduced fear of social disapproval, and being less aware of others can result in ignoring situational cues that allow the powerful to dismiss or ignore even strong social norms in their treatment of the powerless Van Kleef et al., 2008).

The legitimately powerless
Contrarily to the powerful in legitimate hierarchies, the powerless most often hold the same moral standards for themselves as the powerful and rarely engage in hypocritical behavior. The powerless often go so far as to judge their moral misbehaviors more harshly than the moral misbehaviors of the powerful (Lammers et al., 2010). At the same time, power disparities often make the powerless hypersensitive to cues about the inequitable distribution of influence and resources (Anderson et al., 2006;Mason et al., 2010). Consequently, the powerless socially constrain themselves to keep in the good graces of the powerful (Handgraaf et al., 2008;Hays & Goldstein, 2015). Therefore, the powerless in legitimate hierarchies conform to the expectations of and give more latitude toward the powerful to reach a desired goal or reward while at the same time avoiding threats and punishment (Hays & Goldstein, 2015). In the classroom, students' perceptions of power impacts their ability to learn (Lovorn et al., 2012). A negative impact of power on students includes (a) taxing emotions such as anxiety and depression, (b) focusing on threat and punishment in the environment, (c) second-guessing themselves, and (d) social withdrawal (Sharabi-Levine, 2016). Students can also cope by idealizing professors and not exercising the power that they do have to complain when power dynamics are abused. Often being in a legitimate hierarchy makes those without power to believe that they are not entitled to exercise any power (Heator, 2018;Lammers et al., 2010). Students understand that they need to complete training requirements to complete their degree. They often feel obligated to not report abuses of power by professors in order to ensure timely degree completion as well as ensure that they are not additionally targeted for reporting abuses of power (Cruikshanks & Burns, 2016).
Further compounding the potential exploitation of the powerless, researchers suggest that the powerful and the powerless have an extremely difficult time taking each other's perspectives (Heator, 2018). This occurs due to the overwhelming bias that exists through the differing perspectives taken simply by being a member of the powerful or powerless group, which therefore makes the person unable to completely comprehend the perspective of a member of the opposing group (Handgraaf et al., 2008).

Boundary-crossings in counselor education training programs
The previous section has explained how power works in all legitimate hierarchies. The focus now shifts to how legitimate hierarchies function in counselor education programs. CEs hold legitimate power over CESs through having the authority to make rules, enforce those rules, as well as preserve moral norms by judging the actions of CESs (Bolkan & Goodboy, 2013;Goodboy et al., 2011). CEs influence CESs in several ways: grading assignments, writing recommendation letters, providing feedback for counseling skill development, research mentoring, supervision of clinical skills, and gate-keeping entrance to the profession (Gu et al., 2011). CEs are guided in their conceptualizations of boundary-crossings with CESs by the standards defined in the American Counseling Association Code of Ethics (ACA, 2014) and the Association for Counselor Education and Supervision (ACES) Standards for Counseling Supervisors (ACES Supervision Interest Network, 1990). In addition to codes and standards, experts suggest that ethical decision-making models assist with justifiable ethical decision-making (Gottlieb et al., 2007;Kennedy, 2008;Kitchener & Anderson, 2010). As such, CEs leave their doctoral training programs with training in these standards as well as the use of ethical information resources when negotiating boundary-crossings. Likely, most CEs leave their training programs feeling prepared to negotiate boundary-crossings successfully with CESs.
When negotiating boundary-crossings, CEs and CESs can use the ACA (2014) Code of Ethics, Gottlieb's (1993) Decision-Making Model, and Burian and Slimp (2000) Social Dual-Role Relationship Model as ethical decision-making resources as they all specifically assist in navigating boundary crossings. Section F, Supervision, Training, and Teaching of the ACA (2014) Code of Ethics (p. 18) assists CEs and CESs in recognizing the professional obligations that CEs hold towards CESs (Jungers & Gregoire, 2016). Increased awareness of Section F can guide CEs and CESs when navigating boundary-crossings between the two groups. Gottlieb's Decision-Making Model helps negotiate boundary-crossings in counseling settings between groups with and without power (Barnett et al., 2007;Kitchener & Anderson, 2010). Gottlieb's model guides the user in three ways to evaluate power differentials: (a) the current role, (b) the contemplated role, and (c) if increasing power disparities in the contemplated role could negatively impact the current role. Next, the user considers the length of the current relationship and if the person without power can end the relationship without negative consequences. Lastly, Burian and Slimp's model has the user contemplate motives when evaluating boundary-crossings specifically between CEs and CESs. Burian and Slimp's model first suggests that the user classify the boundary-crossing into one of four categories: (a) the highest priority, the professional needs of CESs; (b) the second-highest priority, the professional needs of CEs; (c) the third highest priority, the personal needs of CESs; and (d) the last priority, the personal needs of CEs. Boundary-crossings that fall into the first category of meeting the professional needs of CESs are evaluated as likely acceptable. However, Burian and Slimp caution that boundary-crossings categorized in the second-highest priority hold greater concern than the first category. The third category holds a much greater risk of becoming harmful than the second category and that the last category should not be considered as counselor educators should not look to students to fulfill their personal needs. As the categorization of the boundary-crossing moves from the first to the last category, the risk of potential harm to CESs increases and warrants special consideration. Burns and Cruikshanks (2019) studied CEs' use of ethical information resources when evaluating 16 boundary-crossings with CESs (see Table 1). The researchers reported that out of the four conditions in the study, Burian and Slimp's model was the only ethical decision-making resource to change CEs views on boundary-crossings, which could mitigate potential harm to CESs. Burian and Slimp's model significantly changed CEs' perceptions of six boundary-crossings to be less ethical. Surprisingly, almost 60% of CEs reviewing Burian and Slimp's model reported they would not consider using the model in the future because they disagreed with the model's results and because it categorized boundaries as black and white. Burns (2019) studied CESs' use of ethical information resources when evaluating 16 boundary-crossings with CEs. CESs were more concerned when the boundary-crossing impacted their personal needs. Ultimately, the ACA Code of Ethics only influenced CESs to view one boundary-crossing as significantly less ethical as compared to the pre-test. Burian and Slimp's model only influenced CESs to view two boundary-crossing as significantly less ethical as compared to the pre-test. Only one-third of CESs negatively experienced using Burian and Slimp's and Gottlieb's models stating that the models were complicated, restrictive, and missed nuances.

Purpose
Counselor education training programs require the constant negotiation of boundary-crossings between CEs and CESs. Based upon the literature review of power differentials in legitimate hierarchies, CEs and CESs likely hold different views of boundary-crossings between CEs and CESs. Further, it is likely that CESs overly forgive and comply even when they are uncomfortable with boundary-crossings between CEs and CESs. It would be helpful for CEs to understand how power differentials influence boundary-crossing perceptions within the legitimate hierarchy of counselor education. Burns and Cruikshanks (2019) and Burns (2019) used the same 16 boundarycrossing scenarios and the same four ethical-information resources with CEs and CESs to understand how these resources influenced each group when contemplating boundary-crossings with the other group. For this study, I used the ex-post facto data from both prior studies to test differences in original reactions on the 16 boundary-crossing scenarios between CEs and CESs.

Table 1. Pre-test T-tests for counselor educator (CE) and counselor education student (CES) groups CE Pre-test#
The counselor educator tells a student that they also have an Autistic child and then discusses the community resources available for Autistic children with the student 5.08 .13 14. The counselor educator has a student write an IRB proposal, collect and analyze data, and submit a journal article for review. The counselor educator receives first author status on the publication Additionally, I used the ex-post facto data from both studies to test differences between CEs and CESs in the ratings of 16 boundary-crossing scenarios after reviewing one of the four ethicalinformation resources. Lastly, I used the ex-post facto data from both studies to test pre-test to post-test differences between CEs and CESs in the ratings of 16 boundary-crossing scenarios after reviewing one of the four ethical-information resources.
Based on previous research on power in legitimate hierarchies, I hypothesized that CESs would initially rate the 16 boundary-crossing scenarios as more unethical compared to CEs. Additionally, I hypothesized that reviewing the ACA Code of Ethics, Gottlieb's model, or Burian and Slimp's model would influence CESs to rate the 16 boundary-crossing scenarios as more unethical as compared to CEs. Lastly, I hypothesized that the ACA Code of Ethics, Gottlieb's model, and Burian and Slimp's model would influence CEs and CESs to rate the 16 boundary-crossing scenarios as more unethical when comparing their views to the pre-test. Burns and Cruikshanks (2019) stated counselor educators worked full-time in a master's graduate program teaching courses specific to addiction counseling, career counseling, clinical mental health counseling, clinical rehabilitation counseling, college counseling, marriage couple and family counseling, and school counseling. Burns (2019) stated counselor education students were individuals actively taking courses in a CACREP accredited master's graduate program. Burns and Cruikshanks (2019) and Burns (2019) performed power analyses with G*Power using a medium effect size of.25, power level of .80, and alpha of .05 (Faul et al., 2007), which recommended a minimum sample size of 136 for each study. Burns and Cruikshanks (2019) and Burns (2019) selected a medium effect size since average variance was expected and important, practical concerns for the counseling profession were desired (Healey & Hays, 2012). (2019) stated counselor educator (CE) participants (n = 207) averaged 45 years of age (Range = 29-75, SD = 10.7). Most participants identified as female (n = 131, 63%), graduating from a CACREP accredited program (n = 129, 62%) and European American (n = 152, 73%). CEs taught in several academic settings: public college or university with master's counselor education programs (n = 90, 44%), public college or university with master's and doctoral counselor education programs (n = 71, 34%), private college or university with master's counselor education programs (n = 30, 14%), and private college or university with master's and doctoral counselor education programs (n = 16, 8%).

Ethical information resources
Ethical information resources in both studies were defined as resources a CE or CES could use when considering the ethical implications of boundary-crossings with a CE or CES (Burns, 2019;Burns & Cruikshanks, 2019). The data analyses in this study were generated from the raw data of two separate studies (Burns, 2019; Burns & Cruikshanks, 2019) and therefore used the same ethical information resources: (a) the ACA (2014) Code of Ethics, (b) Gottlieb's (1993) Decision-Making Model, and (c) Burian and Slimp (2000) Social Dual-Role Relationship Model. As a control group, a placebo (d) magazine article was also used.
The previous studies used Gottlieb's (1993) Decision-Making Model and Burian and Slimp (2000) Social Dual-Role Relationship Model due to their exclusive focus on ethical decision-making when negotiating boundary-crossings in counseling settings. Participants in both studies reviewed Gottlieb's model and Burian and Slimp's model through verbatim portions of text and one figure from each model. The fourth group received a placebo and read the same magazine article: Ethical Epitaphs: Extended Explorations of Excellence (Handelsman, 2010). The placebo groups' magazine article required the same amount of time from participants as the rest of the ethical information resources but did not provide guidance on ethical decision-making. Participants received complete reference details for the four resources. Burns and Cruikshanks (2019) used a list of 1,624 CEs from all 50 states teaching in addiction counseling, career counseling, clinical mental health counseling, clinical rehabilitation counseling, college counseling, marriage couple and family counseling, school counseling, and counselor education and supervision counselor education programs. A random sample of 850 CEs were contacted, and 207 participants completed the study, which resulted in a 24% response rate. Burns (2019) specifically recruited CESs from six different CACREP programs: one public college with only a master's program, one public college with a masters and doctoral program, one private college with only a master's program, one private college with a masters and doctoral program, one for-profit college with only a master's program, and one for-profit college with a master's and doctoral program. Approximately 1,290 CESs were contacted, and 224 completed the study, which resulted in a 17% response rate.

Data collection procedures
In both studies, participants received an e-mail that included a link to a webpage that contained the informed consent form, a demographic questionnaire, and all study materials. Study participants were randomly assigned to one of the four ethical information resources. First, participants gave consent, answered demographic questions, and rated the 16 boundary-crossing scenarios without any ethical information resource. Next, they saw one of the four ethical information resources (the ACA (2014) Code of Ethics, Gottlieb's (1993) Decision-Making Model, Burian and Slimp (2000) Social Dual-Role Relationship Model or the placebo magazine article) and then rerated the initial 16 boundary-crossing scenarios. Lastly, they rated the one ethical information resource that they were provided in the study and had the option of anonymously providing their name and e-mail address to receive an e-gift card. CEs spent a median time of 19 minutes and CESs spent a median time of 32 minutes completing the study. This time does not include filling out gift-card information. The longer time needed to complete the study may account for the lower response rate.

Measure
The instrument used in both studies (Burns, 2019;Burns & Cruikshanks, 2019) was based upon higher education students' identification of five types of interactions occurring between faculty and students: sexual/dating, friendship/social, personal/counseling, business/financial, and professional/academic (Hoppe, 2013). Additionally, the instrument was shaped by four-year colleges' and universities policies' in the United States, the American Association of University Professors (2009,2015)) policies, the 2014 ACA Code of Ethics, existing research on student reports of boundarycrossings with educators, and Burian and Slimp (2000) suggestion that relationship motives are important factors when considering boundary-crossings. As such, the 16 scenarios (see Table 1) were evenly distributed among Burian and Slimp's model stating that the CE should classify the boundary-crossing into one of four categories: (a) the highest priority, the professional needs of CESs; (b) the second-highest priority, the professional needs of CEs; (c) the third highest priority, the personal needs of CESs; and (d) the last priority, the personal needs of CEs. Boundary-crossings that fall into the first category of meeting the professional needs of CESs are evaluated as likely acceptable. Additionally, the 16 scenarios between CEs and CESs ranged from being likely ethical given certain contexts to being unethical in all contexts. The 16 scenarios (Table 1) were brief to measure the general opinion of each potential boundary-crossing.
Participants used a Likert scale for the 16 boundary-crossing scenarios: responding highly unethical received a value of 1 and responding highly ethical received a value of 6. Participants rated the resource's helpfulness: responding highly helpful received a value of 1 and responding highly unhelpful received a value of 6. Afterward, participants rated the likelihood of using the resource in the future: responding highly likely received a value of 1 and responding highly unlikely received a value of 6. Burns and Cruikshanks (2019) and Burns (2019) checked the data in both studies for outliers by running frequency reports and found no evidence of miscoded data or univariate outliers. Descriptive statistics were obtained from Burns and Cruikshanks (2019) and Burns (2019) to report participant demographics and ratings of the 16 boundary-crossing scenarios for CEs and CESs. I used independent samples T-tests to determine differences between CEs and CESs on the pretest ratings of the 16 boundary-crossing scenarios. I performed independent samples T-tests to examine post-test differences on the 16 boundary-crossing scenarios for CEs and CESs within each of the four ethical information resource groups. I used Holm's Sequential Bonferroni Procedure with all independent samples T-tests. The Holm's Sequential Bonferroni Procedure was selected because it is more powerful than the single-step Bonferroni and it is more conservative (Cleophas & Zwinderman, 2011).

Results
I ran independent samples T-tests to evaluate differences between CEs' and CESs' pre-test impressions of the 16 boundary-crossing scenarios and found that 11 scenarios were significantly different (Table 1). After using the Holm's Sequential Bonferroni Procedure, six scenarios met the cutoff. CESs found it less ethical at a small magnitude for a CE to hug a CES (t = −4.07, p > .01). CESs found it less ethical at a moderate magnitude for a CE to accept a small gift from a CES (t = −4.50, p > .01), attend a CES's big event (t = −4.46, p > .01), and carpool with a CES to a meeting (t = −5.98, p > .01). CEs found it less ethical at a small magnitude for a CE to have a business partnership with a CES (t = 4.09, p > .01). CEs found it less ethical at a moderate magnitude for a CE to engage in personal counseling with a CES (t = 5.66, p > .01).
I ran independent samples T-tests to evaluate changes between CEs' and CESs' post-test impressions of the 16 boundary-crossing scenarios by ethical information resource (Table 2). I found seven significant differences between CEs and CESs in the placebo groups. After using the Holm's Sequential Bonferroni Procedure four scenarios met the cutoff. CESs found it less ethical at a moderate magnitude for a CE to accept a small gift from a CES (t = −2.44, p > .01), attend a CES's big event (t = −2.52, p > .01), and "friend" a CES on Facebook (t = −2.56, p > .01). CEs found it less ethical at a moderate magnitude for a CE to engage in personal counseling with a CES (t = 2.56, p > .01). I found five significant differences between CEs and CESs in the ACA code groups. After using the Holm's Sequential Bonferroni Procedure four scenarios met the cutoff. CESs found it less ethical at a moderate magnitude for a CE to hug a CES (t = −3.96, p > .01). CEs found it less ethical at a moderate magnitude for a CE to have a business partnership with a CES (t = 3.81, p > .01) and engage in personal counseling with a CES (t = 4.01, p > .01). CESs found the ACA Code of Ethics more helpful to their boundary-crossing decision-making than CEs at a moderate magnitude (t = −3.38, p > .01).
I found seven significant differences between CEs and CESs in the Gottlieb groups. After using the Holm's Sequential Bonferroni Procedure, seven scenarios met the cutoff. CESs found it less ethical at a moderate magnitude for a CE to accept a small gift from a CES (t = −3.36, p > .01), attend a CES's big event (t = −3.48, p > .01), and hug a CES (t = −3.38, p > .01). CEs found it less ethical at a moderate magnitude for a CE to not meet with a CES for required supervision (t = 3.71, p > .01) and engage in personal counseling with a CES (t = 3.64, p > .01). CESs reported Gottlieb's model was more helpful (t = −3.41, p > .01) and more likely to be used in the future (t = −3.33, p > .01) when negotiating boundary-crossings than CEs at a moderate magnitude. I found eight significant differences between CEs and CESs in the Burian and Slimp groups. After using the Holm's Sequential Bonferroni Procedure, six scenarios met the cutoff. CESs found it less ethical at a moderate magnitude for a CE to carpool with a CES to a meeting (t = −3.24, p > .01). CEs found it less ethical at a small magnitude for a CE to take first author status when the CES performed all the work (t = 3.51, p > .01). CEs found it less ethical at a moderate magnitude for a CE not to meet with a CES for required supervision (t = 2.98, p > .01), have a business partnership with a CES (t = 3.14, p > .01), engage in personal counseling with a CES (t = 3.47, p > .01), and not provide a CES with comprehensive feedback because of the CE's feelings towards the CES (t = 3.32, p > .01). CESs found Burian and Slimp's model more helpful to their boundary-crossing decision-making than CEs at a small magnitude (t = −1.93, p = .05)

Discussion
The results of this study demonstrate important differences between CEs and CESs on the pre-test and post-test viewpoints. On the pre-test, CEs and CESs expressed significantly different views on six boundary-crossings. Overall, pre-test differences demonstrated that CESs perceived boundarycrossings differently due to power differentials (Anderson et al., 2006;Mason et al., 2010) when their personal needs could be impacted in intimate ways (small gifts, attending a big event, carpooling to meeting, and hugging a student) (Burian & Slimp, 2000). CEs were more concerned about boundary-crossings when CESs' personal needs could be impacted by non-mandated dual role relationships (business partnerships with students, and performing personal counseling with students) (Burian & Slimp, 2000).

Post-test differences by group
It is noteworthy that CEs and CESs were influenced in a similar direction for three out of the four significant post-test boundary-crossings in the placebo group. CEs and CESs both shifted to a more ethical rating for attending a big event and providing personal counseling. CEs and CESs both shifted to a less ethical rating towards accepting a small gift. CEs shifted to a more ethical and CESs to a more unethical rating towards "friending" on Facebook. After reading the magazine article, the increased disparity between CEs and CESs for "friending" on Facebook resulted in this boundary-crossing being significantly different for the first time. The magazine article influenced CEs and CESs to reconsider four boundary-crossings that impacted CESs' personal needs. As the magazine article offered no help in ethical decision-making, it comes as no surprise that both groups rated current its helpfulness and future use similarly.
CEs and CESs were influenced in a similar direction by the ACA Code of Ethics for two out of the three significant post-test boundary-crossings. CEs and CESs shifted to a less ethical rating for providing personal counseling to a CES and hugging a CES. CEs shifted to a less ethical, and CESs shifted to a more ethical rating for entering into a business partnership with a CES. The ACA Code of Ethics influenced CEs and CESs to reconsider three boundary-crossings that impacted CESs' personal needs. As CEs developmentally have tremendous exposure to the ACA Code of Ethics as compared to CESs, it is not surprising that CESs would find the ACA Code of Ethics more helpful than CEs at a medium effect size. CEs and CESs similarly rated use of the ACA Code of Ethics in the future between highly and moderately likely. Since the ACA Code of Ethics plays such a prominent role in the counseling profession, it is helpful to understand that CEs and CESs share a similar appreciation of the resource.
CEs and CESs were influenced in a similar direction by Gottlieb's model for four out of the five significant post-test boundary-crossings. CEs and CESs shifted to a less ethical rating for taking small gifts, attending a big event, hugging, and providing personal counseling. CEs shifted to a less ethical rating, and CESs obtained the same rating for not providing the required supervision. After reviewing Gottlieb's model, CEs shifted to a more unethical stance, which increased the disparity between CEs and CESs and resulted in this boundary-crossing being significantly different for the first time. Gottlieb's model influenced CEs and CESs to reconsider four boundary-crossings that impacted CESs' personal needs and one boundary-crossing that impacted CESs' professional needs. CESs rated the helpfulness of Gottlieb's model and the future use of Gottlieb's model significantly higher at a medium effect size as compared to CEs. CEs likely felt less influenced by Gottlieb's model due to past experiences with boundary-crossing decision-making and may already prefer a different ethical decision-making model (Burns & Cruikshanks, 2019). Because CESs have had less exposure to both the use of ethical decision-making models as well as boundary-crossings between CEs and CESs, it is not surprising that CESs would find Gottlieb's model significantly more helpful at a medium effect size as compared to CEs.
CEs and CESs were influenced in a similar direction by Burian and Slimp's model for three out of the six significant post-test boundary-crossings. CEs and CESs shifted to a less ethical rating for carpooling, business partnerships, and personal counseling. CEs shifted to a less ethical rating and CESs to a more ethical rating for not meeting for required supervision, not providing comprehensive feedback, and inappropriately claiming first author status. After exposure to Burian and Slimp's model, the increased disparity between CEs and CESs resulted in these last three boundarycrossings being significantly different for the first time. Burian and Slimp's model influenced CEs and CESs to reconsider three boundary-crossings that impact CESs' personal needs and three boundary crossings that impacted CESs' professional needs. Regarding helpfulness, CESs rated the helpfulness of Burian and Slimp's model significantly higher at a small effect as compared to CEs. As Burian and Slimp's model is specific to counselor education and supervision, it was surprising that CEs did not rate the future use of Burian and Slimp's Model's significantly higher than CESs since CEs hold the burden of ensuring the protection of CESs.

Limitations
Several limitations should be considered when reviewing this research. The two separate cross-sectional studies captured the views of participants at one point in time. There was no consideration of CEs or CESs developmental stage. If other boundary-crossings and ethical information resources were used in previous studies, other outcomes might be found. Only certain types of CEs and CESs may have volunteered to participate based on the topic of the study. CESs from CACREP accredited programs may have different views from CESs in non-CACREP accredited programs. CESs who had experienced a boundary-crossing with a CE may have avoided participating in the study as the researcher is a counselor educator. As participants were asked ethics questions, they may have used socially desirable responding. The results of this study may be skewed since 84% of CES participants reported never experiencing a boundary-crossing with a CE. Lastly, as CEs were recruited from CACREP and non-CACREP accredited counselor education programs, this research does not entirely reflect CEs graduating from CACREP accredited programs.

Suggestions for future research
The study should be replicated to verify the findings. Researchers could explore alternative boundary-crossings and ethical information resources. More research is needed to determine if past boundary-crossing concerns, age, gender, race, and other aspects of diversity impact how CEs and CESs use ethical information resources when making boundary-crossing decisions. Finally, qualitative studies could further explore the lived experiences of CEs and CESs trying to negotiate boundary-crossings that meet the professional and personal needs of CEs and CESs.

Implications for practice and training
The vast majority of CESs (84%) participating in this study denied ever experiencing a boundarycrossing with a CE. CEs should celebrate this statistic. Because CESs do not report experiencing boundary-crossings with CEs, this reduces the chances that there is a negative impact of power on students. It is unlikely that CESs are experiencing (a) taxing emotions such as anxiety and depression, (b) focusing on threat and punishment in the environment, (c) second-guessing themselves, and (d) social withdrawal (Sharabi-Levine, 2016). The lack of reports of boundary-crossings with CEs by CESs may suggest that CEs operate in this legitimate hierarchy with a low level of entitlement, hypocritical, and exploitation behaviors (Lammers et al., 2010;Magee & Smith, 2013;Sharabi-Levine, 2016;Smith & Bargh, 2008). At the same time, it is possible that the lack of reports by CESs of boundary-crossings with CEs could also be a product of CESs giving more latitude to CEs, being conforming to CEs, and idealizing CEs (Hays & Goldstein, 2015;Lammers et al., 2010).
Ultimately, a professor's position of authority and possession of expertise affords them the power to asymmetrically influence students and distribute resources (Bolkan & Goodboy, 2013;Goodboy et al., 2011). As such, it is important for CEs to understand how CESs perceive boundary-crossings in counselor education because it is often challenging for the powerful to see the powerless' views completely (Heator, 2018). As in all legitimate hierarchies, those in power often believe that they do not abuse their power (Lammers et al., 2010). It is likely that CEs hear few complaints about boundary-crossings from CESs and feel confident in their ability to identify and successfully negotiate boundary crossings with CESs. Still, it would be helpful for CEs to understand that CESs were very concerned in this study about boundary-crossings that encroached upon their personal needs that were not a requirement for successfully graduating from a counselor education program. Consequently, CEs may want to become clearer in their understanding of the differences between optional behaviors versus those of mandated requirements for graduation. A CE and a CES sharing a car ride to a professional meeting is an optional boundary-crossing of preference that does not assist a CES in completing the mandated requirements for graduation. This contrasts with the mandated requirement that a CE provide the required supervision hours to a CES.
Additionally, it might be helpful for CEs to understand that CESs are less concerned about boundary-crossings that specifically focus on CEs not performing their professional obligations to CESs. From their developmental perspective, CESs cannot completely understand the exact professional obligations that CEs are required to offer CESs nor the importance of upholding those professional obligations to meet CESs' professional needs. CESs are more likely to view a CE relaxing or ignoring professional requirements to successfully graduate from a counselor education program as a desired outcome. CESs, at that moment, will likely appreciate reducing the accountability requirements of the academic curriculum. As stated by Lammers et al. (2010), it is the total obligation of the powerful to establish and maintain their obligations to the powerless.
This study suggests that CEs and CESs place different weights on boundary-crossings between the two groups. When CEs are negotiating boundary-crossings with CESs, this study suggests that Burian and Slimp's model would likely be most helpful as it is specific to counselor education and requires CEs to contemplate motives. Burian and Slimp's model first suggests that the user classify the boundary-crossing into one of four categories: (a) the highest priority, the professional needs of CESs; (b) the second-highest priority, the professional needs of CEs; (c) the third highest priority, the personal needs of CESs; and (d) the last priority, the personal needs of CEs. However, it was concerning that almost 60% of CEs reviewing Burian and Slimp's model reported they would not consider using the model in the future. It could be that CEs feel confident that boundary-crossing issues are not a concern in their counselor education programs. CEs may feel that they are already well trained to be aware of and therefore avoid boundary-crossing concerns with students. At the same time, CEs hold legitimate power over CESs regarding requirements for successfully graduating from a counselor education program (Hays & Goldstein, 2015). As outlined in the literature review, those with power in legitimate hierarchies have the potential to be unaware of how their power can negatively impact those without power Magee & Smith, 2013;Sharabi-Levine, 2016). Burian and Slimp's model in this study did the most to raise awareness of boundary-crossings that could negatively impact CESs' personal needs and professional needs. As Gottlieb's global counseling boundary-crossing model came in second in this study, it is possible that any ethical decision-making model would likely raise awareness more than not using an ethical decision-making model. While it is always wise for CEs to review the ACA Code of Ethics, they should be aware that they might feel less cautious about boundary-crossings than before referring to the ACA Code of Ethics.
However, when CESs are negotiating boundary-crossing concerns with CEs, the results suggest that Gottlieb's model would likely be most helpful. As Burian and Slimp's specific CE to CES model came in second, it is possible that any ethical decision-making model would likely raise awareness in CESs more than not using an ethical decision-making model. While it is always wise for CESs to review the ACA Code of Ethics, they should be aware that after reviewing the document that they could give CEs more latitude in boundary-crossings.
The results of this study suggest that the profession of counseling has much to celebrate in terms of the negotiation of boundary-crossings between CEs and CESs. At the same time, the results of this study support previous findings that the powerless (CESs) give latitude and deference to the powerful (CEs) (Hays & Goldstein, 2015) even after reviewing ethical decision-making resources. Additionally, the legitimately powerful (CEs) feel confident in their understanding of acceptable and unacceptable behavior (Lammers et al., 2010). CEs influence CESs in grading assignments, supervising clinical skill development, and gate-keeping entrance to the profession (Gu et al., 2011). As such, it is important to be aware of effective ethical decision-making resources that can reduce any potential for harm that can creep into any legitimate hierarchy Van Kleef et al., 2008).