Housing Pros

Research Results: Readiness of Resident Assistants to Manage Suicidal Crises

By Jayoung Choi, M.S.S.A, Cynthia A. Yamokoski-Maynhart, M.A., & James R. Rogers Jr., Ph.D., The University of Akron

Up until recently, the law has seemed to protect higher learning institutions against legal responsibility for student deaths by suicide; however, recent legal trends suggest that this protection may be diminishing (Lake & Tribbensee, 2002). For example, in 2000 Ferrum College in Virginia was legally found to have “shared responsibility” for the suicide of a student that occurred in the campus dorm.

Given the fact that suicide is the third leading cause of death among the US college-aged population, 18 to 24 year olds (Barrios, Everette, Simon, & Brenner, 2000), such a trend poses a concern for colleges. Although the suicide rate for college students is difficult to determine accurately, it is generally reported to be lower than that of the same age group in the general population. Based on their longitudinal study of Midwestern university campuses Silverman, Meyer, Sloane, Raffel, and Pratt (1997) reported a rate of 7.5/100,000 for college students. This rate was approximately one half of the suicide rate for a sample matched by age, gender, and race. However, beyond completed suicide, data also suggest that suicidal ideation is prevalent in the college population. For example, Furr, Westefeld, McConnell and Jenkins (2001) reported that 32% of a sample of 1,455 college students experienced thoughts of suicide and 1% had made a suicide attempt while attending college. Again, among 4,838 college student who participated in the 1995 National College Health Risk Behavior Survey (NCHRBS), 11.4% reported having experienced suicidal ideation during the preceding 12 month period and 1.7% attempted suicide (Barrios et al., 2000). Furthermore, Barrios and her colleagues suggested the first and second leading causes of death, unintentional and homicide, among this group may also link to suicide ideation. Their data showed that college students who reported suicide ideation were significantly more likely to engage in other high risk behaviors such as riding with drivers who had been drinking, drinking alcohol themselves, or carrying a firearm. Given the high prevalence of suicide ideation and the potential link to other high risk behaviors, an effective, comprehensive suicide prevention and intervention training in college campus seems paramount.

Resident assistants (RA’s) are often the first line of defense for suicide prevention and intervention, but little is known about the training, efficacy, and effectiveness of staff in managing suicidal crises. In outlining guidelines of training RA’s as interventionists, Grosz (1990) emphasized the importance of the RA as “a key suicide prevention member” (p193) in dealing with suicidal students and suggested six critical components in training RA’s: 1) campus policy and procedure regarding suicide and suicidal behavior, 2) personal feelings about suicide, 3) recognition of the suicidal student, 4) screening and assessment, 5) the referral process, 6) dealing with aftermath of suicide (see Grosz, 1990 for detailed information). Lake and Tribbensee (2002) also argued that effective suicide prevention requires focused training for all campus staff including resident assistants. For example, in the case of suicide by a freshman in the University of Iowa who killed himself in his dorm, the Iowa Supreme Court dismissed the case for legal duty of the university by largely relying on the fact that the resident assistant and the assistant director for resident life were deemed to have managed the situation with reasonable care by offering support and continuing to encouraging the student to seek counseling (Lake & Tribbensee, 2002).

The purpose of this study, therefore, was to examine RA’s’ training experience and preparedness in managing suicidal crises and to offer recommendations for future training. Our research questions were: how much and what type of training experiences have residential staff received, how efficacious do staff feel in handling crises, and how skilled are resident assistants in intervening with suicidal students? The relations between these variables were analyzed to determine the needs of RA’s, the specific training variables that are related to their sense of efficacy, and effectiveness in their role as crisis interventionist.

Method

Participants

Participants in the study included 100 resident assistants in colleges and universities nationwide. The average age of participants was 21.04 (SD = 3.38), 29.5% were male and 70.5% female, 63.2% were either sophomores or juniors, over half were in their first year as an RA, and their ethnicity was similar to national averages (Table 1). Participants also shared information about the schools where they were RA’s. Most described their school as large, public, urban and commuter schools.

Procedure

Data was collected through both an internet survey run by Survey Monkey (n = 63) and paper and pencil administration at a public urban university (n = 37). Internet participants were recruited over the Reslife.Net website, and those who were interested were directed to the anonymous online survey. The participants using paper pencil administration from one university were provided with copies of the measure and consent form, and were asked to return survey via mail if they were interested in participating. The response rate for the paper and pencil administration was 51%.

Measures

Intervention competence/actual abilities. This variable will be measured using the Suicide Intervention Response Inventory-2 (SIRI-2: Neimeyer & Bonnelle, 1997). This scale is designed to measure the intervention skills of paraprofessionals (i.e., counselors, hotline volunteers, teachers) when responding to suicidal individuals. The scale consists of 25 client statements and two possible helper responses to the statement. For each statement, respondents are asked to rate the appropriateness of the response from –3 (highly inappropriate response) to +3 (highly appropriate response). The scale is scored by calculating summed difference scores between each individual’s responses and the average rating of a group of suicide experts. Therefore, the higher score on SIRI-2 means the less skillful the respondent is assumed to be in responding to a distressed individual. Cut-off scores have not yet been established. The SIRI-2 was updated from the original SIRI (Neimeyer & MacInnes, 1981) to better discriminate abilities, but the two are related to each other, so reliability and validity evidence is based upon both the original and updated versions. Reliability has been demonstrated for both the original version and the SIRI-2 (a = .78-.93; two-week test-retest r = .79-.92). The internal consistency reliability of the SIRI-2 was .79 for this study. Construct-related validity of SIRI-2 was supported by the significant difference in scores between masters-level counselors and introduction to psychology students, and scores improved following education and training in suicide intervention skills (Neimeyer & Bonnelle, 1997). Convergent validity was assessed by examining the relations between the original SIRI and other measures of counseling skills (Neimeyer & MacInnes, 1981). A moderate correlation revealed that the SIRI is related to counseling skills and training, however, this scale provides additional and unique information over and above these other skills (Neimyer & Bonnelle, 1997).

Preparation. RA’s level of preparation and their perceived preparedness was measured using questions created by the experimenters. Included were questions related to various training activities, the total time spent in training, their own perceptions of their level of preparation, and their desire for further training experiences. The format for these questions included Likert, open questions, and yes/no questions.

Efficacy/Perceived effectiveness. This variable was measured using a Likert scaled question created by the researchers for the purpose of this study. A scale assessing the efficacy of managing suicidal crises could not be located

Results and Discussion

What are the training experiences of resident assistants?

On average, RA’s said that they received 5.69 hours (SD = 10.19) of training on suicide, and nearly all of the RA’s (95.3%) reported that they received either voluntary or mandatory training regarding working with suicidal residents. The vast majority of RA’s had received training on all of the topics presented (percentages ranged from 62.5% to 91.5%), with the exception of working with families of students in crisis (only 21.8% received training on this topic) (Table 2). Three-fourths of the RA’s said that they received training on how to manage a suicidal crisis. While this is clearly the majority of students, one-fourth of all RA’s surveyed did not receive such training, and over half of RA’s (54.8%) said that they had encountered a suicidal resident. Among 48 respondents who had encountered at least one suicidal resident, two had never received any training. Training was most often received from resident life staff (92.9%), and over half of the RA’s (60.5%) said that they acquired additional training from reading on their own.

Even though most RA’s received training on all the topics concerning working with suicidal residents, a large majority desired further training on all of these topics. The percentage of RA’s desiring further training across the various topics ranged from 68.1% to 81.7%.

What factors are related to and predict perceived and actual preparedness and effectiveness?

Participants’ scores on the SIRI-2 reflect their competence in the skills required for suicide crisis intervention. Those with previous encounters with suicidal residents did not demonstrate different levels of competence than those who had never had this experience when SIRI-2 scores were compared using one-way ANOVA [F(1,67) = 3.70, p = .06]. In other words, competence was no different for those with experience with suicidal individuals than those without similar experiences. However, the amount of experience of RA’s was somewhat limited (range = 0-8, mean = 1.44, mode = 0), so perhaps this small amount of experience was not enough to enhance their skills and competence levels. In addition, the correlation between the number of suicidal residents RA’s encountered and their competence was not significant (r = -.20, p = .11). Because of the extreme nonnormality of the training hour variable, we performed Spearman rank order correlations to examine the relationship between hours in training and SIRI-2 scores. Similar to the fact that direct experience with suicidal residents was not related to competence and those with experience did not demonstrate higher levels of competence, hours of training were not significantly related to competence through spearman rank ordered correlation analyses (r=.084, p=.496). So it appears that the training that RA’s received neither enhanced nor hindered the actual competence of RA’s to work with suicidal residents, instead, it appears to not have any relationship. From the information gathered, the relationship between the quality and content of the training experiences and competence cannot be evaluated. Perhaps it is not the amount of hours that is important, as suggested by these results, but instead the content and quality of the training that is more influential.

If RA’s had previous encounters with suicidal residents, their perceived preparedness and effectiveness in managing the situation were examined, and if RA’s did not have previous encounters, their anticipated preparedness and effectiveness if they were to work with a suicidal resident in the future were analyzed. For those with previous encounters, there was no relationship between their actual competence and their perceived preparedness (r = .28, p = .10) and effectiveness (r = .32, p = .06) in managing the crisis. This suggests that RA’s may not be highly accurate in assessing their readiness and effectiveness in working with suicidal individuals. Their perceived preparedness and effectiveness were significantly related to each other (r = .68, p < .001). The number of hours of training was related to RA’s perceived level of preparedness (r=.33, p=.02), but not their perceived effectiveness in managing the suicidal resident (r=.09, p=.57). It may be possible that RA’s are lacking confidence in their skills, and it is unclear how this may impact their actual behaviors if found in a suicidal emergency. Even though RA’s may have felt prepared to deal with suicidal residents based upon the hours they spent in training, it may also be possible that their experience with suicidal residents left them with doubts about whether or not they handled the situation properly, resulting in decreases in their efficacy. Further research is required to test these hypotheses.

For those RA’s without previous encounters with suicidal residents, there again was no relationship between their actual competence and their anticipated preparedness (r = .23, p = .22) and effectiveness (r = .26, p = .16) in working with a future suicidal resident. Their anticipated preparedness and effectiveness were significantly correlated to each other (r = .92, p < .001). The number of hours of training was related to RA’s anticipated level of preparedness (r=.44, p<.01), and their anticipated effectiveness in managing the suicidal resident (r=.35, p<.05). This suggests that the number of hours of training is related to how prepared and able they feel to manage a potential crisis, but their actual competence is not related to their anticipated preparedness and effectiveness. So RA’s may be evaluating their abilities on the amount of time spent in training instead of their actual skills. Again, it is not clear whether it is the quantity of training or the quality and specific experiences that are more important to RA’s.

Limitations of the study

One limitation of this study was that participants were self-selected and there is no means to determine if those choosing to participate are different than those who decided not to participate. A second limitation is the quality or content of training experiences was not assessed, only the quantity. An additional limitation of this study was actual abilities were not evaluated, only the competence levels of RA’s through an analog design. Participants’ responses may be entirely different when responding hypothetically to the brief client remarks on the SIRI-2 than their responses may be when actually face-to-face with a suicidal resident. During the actual crisis, RA’s would be faced with many likely emotional reactions and a higher level of arousal that may impact their responses.

Recommendations of training programs

The hours of training that RA’s received were not related to their actual abilities as measured by the SIRI-2. This could mean that the training that is being provided is not effective and/or the training provided does not focus upon the skills that are measured by the SIRI-2, which are effective and empathetic communication skills. We recommended that RA training include more than an emphasis on knowledge and facts about risk factors and policies/procedures, and include a focus on the actual interpersonal skills that are required to intervene with a student in crisis. Some of these specific skills include reflecting feelings instead of de-emphasizing emotions. These skills will be useful when encountering any resident in need, not just the suicidal student.

There was a discrepancy between RA’s actual competence and their perceived effectiveness in this sample. Based upon this consistent finding, we propose several recommendations. First, training should focus both upon learning the skills required to manage a suicidal crisis (this would increase their competence), and upon the emotional elements of working with suicidal individuals. For example, training could focus upon RA’s worries and concerns because these reactions may hamper their perceived effectiveness. Grosz (1990) similarly recommended that RA’s be given opportunities to explore personal feelings associated with suicide because working with a suicidal student is likely evoke intense feelings of fear, anger, and helplessness in RA’s. Managing these negative feelings is likely to enhance their self-efficacy of dealing with a crisis. Second, trainees should also receive opportunities to practice and to receive direct feedback on their intervention knowledge and skills to enhance their confidence and so that their perceptions of themselves are more consistent with their actual abilities. Role plays are an excellent way to provide these opportunities to RA’s, they also may assist RA’s in the exploration of their emotional reactions, and will increase the focus in training on intervention skills. Finally, following an encounter with a suicidal resident, the RA involved should meet to debrief with a professional or supervisor to process their emotional reactions and to review the steps taken. Care should be taken to ensure that this meeting is not viewed as punitive, rather this meeting is an opportunity to provide support and to assist the RA in growing and learning. A group debriefing should also be held to provide opportunities for all RA’s to voice their concerns and emotional reactions, but not on evaluating the personal performance of the RA who was involved. These debriefing opportunities will provide the RA who was involved with an opportunity to receive feedback and support about their actual skills, which may result in their actual and perceived abilities being more consistent, and all RA’s will be able to explore their negative emotional reactions.

Table 1.

Demographics

	Mean	SD

Age	21.04	3.38
Students responsible for	51.00	50.57
Hours in suicide training	5.69	10.19
Encounters with suicidal students	1.44	1.94

		Frequency	Percentage

Sex (n=95)
	Male	28	29.5%
	Female	67	70.5%
Race/ethnicity (n=95)
	African American	12	12.6%
	Asian American	1	1.1%
	Biracial	3	3.2%
	Hispanic/Latino	1	1.1%
	Native American	0	0%
	European American	76	80.0%
	Other	2	2.1%
Year in school (n=95)
	First year	6	6.3%
	Sophomore	30	31.6%
	Junior	30	31.6%
	Senior	26	27.4%
	Graduate student	2	2.1%
	Other	1	1.1%
Years experience as RA (n=94)
	One year or less	49	52.1%
	Two years	25	26.6%
	Three years	17	18.1%
	More than three	3	3.2%
Number of undergrads in school (n=94)
	< 1000	7	7.4%
	1000-1999	9	9.6%
	2000-3999	14	14.9%
	4000-5999	11	11.7%
	> 6000	53	56.4%
Type of school (n=95)
	Public	81	85.3%
	Private	14	14.7%
Location of school (n=94)
	Urban	49	52.1%
	Suburban	18	19.1%
	Rural	27	28.7%
Setting of school (n=95)
	Residential	35	36.8%
	Commuter	60	63.2%

Table 2.

Training experiences and desires of residential assistants

	Received training		Desire training
	Y	N	Y	N

Warning signs of suicide (n=82, 61)	86.6%	13.4%	80.3%	19.7%
Suicide risk factors (n=82, 61)	82.9%	17.1%	78.7%	21.3%
Intervening with a distressed student (n=82, 60)	87.8%	12.2%	80.0%	20.0%
Managing a suicidal crisis (n=82, 60)	76.8%	23.2%	81.7%	18.3%
Working with families of students in crisis (n=78, 72)	21.8%	78.2%	68.1%	31.9%
Campus policies and procedures when working with suicidal student (n=82, 60)	78.0%	22.0%	75.0%	25.0%
On-campus treatment options/referrals (n=82, 60)	91.5%	8.5%	68.3%	31.7%
When to consult/involve others in time of crisis (n=82, 60)	87.8%	12.2%	78.3%	21.7%
Dealing with students who may have signs of symptoms of distress/bereavement resulting from the suicide of a fellow student/friend? (n=80, 66)	62.5%	37.5%	72.7%	27.3%

Where and/or how was your training received?		% receiving training

	Voluntary or required training (n=85)	95.3%
	Training by residence life staff (n=85)	92.9%
	Training by college counseling center (n=84)	82.1%
	Training by invited outside professionals (n=81)	44.4%
	Own reading (e.g., internet, books) (n=81)	60.5%

Table 3.

Pearson correlation coefficients

	SIRI	Hours training	Perceived preparedness	Perceived effectiveness	Anticipated preparedness	Anticipated effectiveness
Hours training (n=84)	.08a
Perceived preparedness (n=47)	.28	.33a*
Perceived effectiveness (n=47)	.32	.08a	.68**
Anticipated preparedness (n=34)	.23	.44a**	----	----
Anticipated effectiveness (n=37)	.26	.35a*	----	----	.92**
encounter with suicidal residents (n=84)	-.20	-.03a	.19	.00	----	----

* p < .05 **p < .01
Note. a denotes Spearmen rank ordered correlation. The n-size for these variables differed because only those who have encountered a suicidal resident reported their perceived preparedness and effectiveness, and only those who have not encountered a suicidal resident reported their anticipated preparedness and effectiveness.

References