Inventory Development
 Relationships with Other Constructs

 

The third means of establishing convergent validity is to correlate the TSI total scores with various psychological and organizational measures predicted to be positively or negatively related to occupational stress. The Teacher Stress Inventory has been concurrently related to a number of perceptual, affective, and organizational variables commonly thought to be related to teacher stress. In so doing, the concurrent validity of the TSI stress construct with other constructs is established.

Teacher Stress and Burnout

Some limited evidence exists that teacher stress is significantly related to teacher burnout. It stands to reason that those teachers experiencing the most on-the-job stress would also be those most prone to occupational burnout and that positive correlations would result between the stress and burnout levels encountered in the schools. To determine if this is so, the Maslach Burnout Inventory (MBI) (Maslach & Jackson, 1981) was used in combination with the TSI in a number of unpublished investigations. Based on data collected from a number of teacher samples and the use of Pearson correlations, the MBI subscale Emotional Exhaustion was found to be positively and significantly related to the TSI Total Stress Score: rs = .81, .79, .68, and .63 (ns = 24,266, 72, and 28; all ps = .001). Depersonalization was also significantly related to the Total Stress Score: rs = .71, .46, .65, and .50, respectively (ns = 24, 266, 72, and 28; all ps = .001). Mixed results were found for the relationship between the TSI total score and the MBI subscale (lack of) Personal Accomplishment, however. In this case, the data were reflected for the positively stated personal accomplishment items in the original MBI form, to represent a problem with or lack of on-the-job personal accomplishment. The Total Stress Score of the TSI was correlated with a lack of personal accomplishment at the following level in only one of the four samples: .41 (n = 28; p = .05). Apparently, teacher stress strength levels are significantly related to burnout intensity levels reported by teachers for both emotional exhaustion and depersonalization, but not for the lack of personal accomplishment. In comparison to their low-stressed colleagues, teachers under stress are more likely to feel emotionally "wiped out" and to distance themselves psychologically from their students than they are to feel unaccomplished on the job.

This pattern of high-stress, high-burnout relationships remains consistent when inspecting the correlations between the TSI Total Stress Score with the Total Burnout Score. When the Total Stress Score was correlated with the MBI intensity total score, the following correlations resulted: rs = .73,.74,.64, and .66, respectively (ns = 24, 266, 73, and 28; all ps = .001), representing 41% to 55% shared variance between the strength of stressful events and the intensity of reported burnout. In the case of one sample (sample n = 266), and after controlling for variation attributable to background variables in a multiple regression model, up to 71% of the MBI subscale variation and 59% of the total MBI score variation was accounted for by the TSI Total Stress Score. From these data it is clear that those who very strongly experience stressful events are also those at the greatest risk of burning out on the job.

Teacher Stress and Role Problems

It has long been hypothesized yet only recently researched that teacher burnout may be highly related to the role problems that teachers experience. The research to date indicates that teacher burnout is significantly related to both role conflict and role ambiguity (Crane, 1981; Pierson, 1984; Schwab, 1980). It was hypothesized that similar positive relationships would be found between the strength of teacher stress and the degree to which problems such as role conflict and ambiguity are experienced. Having insufficient information about teaching roles, as measured by the Role Questionnaire (note Schwab, 1980)-- usually termed role ambiguity -- was found to be significantly related to the TSI Total Strength Score in the two samples in which the combination of the variables was investigated: rs = .39 (sample n = 73, p = .001) and .41 (sample n = 28, p = .05). Apparently those teachers experiencing on-the-job ambiguity were also those experiencing the strongest stress.

When teachers have sufficient but conflicting information about how to conduct themselves professionally, they are experiencing role conflict. Role conflict, as measured by the Role Questionnaire, has also been investigated in combination with the TSI with a number of samples: rs = .80 (sample n 266, p .01),.61 (sample n = 73, p = .00 1), and .36 (sample n 28, p trend). Those teachers experiencing the most on-the-job conflict were also those experiencing the strongest stress. Also, and based on the magnitude of the correlations between the stress and role measures, it appears that role conflict is more strongly related to teacher stress than is role ambiguity. In most cases both role ambiguity and role conflict, either separately or in combination, acted as significant predictors of stress strength and stress frequency levels in multiple regression equations.

One additional investigation (Golaszewski, Milstein, Duquette, & London, 1984) explored the relationships between role overload, inadequate role preparation, and role underload and the initial form TSI subscales Emotional Manifestations, Physiological-Fatigue Manifestations, and Biobehavioral Manifestations. As hypothesized, role overload was found to be significantly related to the Emotional Manifestations (r = .55, p = .001), Physiological-Fatigue Manifestations (r .45, p = .001), and Biobehavioral Manifestations (r = .24, p .05) subscales.

Similar relationships were found for role underload -- Emotional Manifestations (r = .24, p = .05), Physiological-Fatigue Manifestations (r = .23, p = .05) -- and inadequate role preparation -- Emotional Manifestations (r = .59, p = .001), Physiological-Fatigue (r = .38, p = .05), and Biobehavioral Manifestations (r = .38, p = .05). Those teachers with too wide or too narrow a role span, or who were not adequately prepared to take on the roles in which they were placed, were those who experienced the strongest stress manifestations.

Teacher Stress and Counseling

Is there a relationship between the presence of on-the-job stress and teachers seeking counseling for work-related problems? It was hypothesized that such a relationship would exist and that those in counseling would report significantly higher levels of teacher stress than those not in counseling. Based on the data reported in Fimian and Krupicka (1987; sample n = 365), it was apparent that the presence of counseling was related to the total strength of stress (r = .20, p = .001). With respect to counseling recipient and nonrecipient differences, those encountering work-related problems and receiving professional counseling for problems at work not only experienced higher levels of Professional Distress, but also experienced significantly stronger stress manifestations than did those not in counseling. Consistently, counseled teachers experienced significantly stronger Emotional, Biobehavioral, and Physiological-Fatigue manifestations than those not receiving counseling for job-related problems.

Teacher Stress and Training Adequacy

Does being adequately prepared to take on the role of "teacher" help minimize teacher stress in the future? This question was asked by comparing the stress strength levels of 721 special education teachers who said that their previous training adequately prepared them for teaching vs. those of 992 teachers who said that they were not adequately prepared. Of these teachers, 42% indicated that they were adequately prepared vs. 58% who indicated they were not (Fimian & Conners, 1987). When the two groups were compared with respect to their stress levels, those not receiving an adequate teacher training background reported significantly stronger stress related to Professional Distress, Discipline and Motivation problems, Emotional Manifestations, and Total Stress than did the adequately trained teachers; those not adequately trained experienced significantly weaker problems related to Personal/Professional Stressors. Apparently, and in most stressful incidents, teachers who felt inadequately prepared were those who experienced the strongest stress in the classroom. Which comes first -- perceptions of inadequate training or on-the-job stress? Which variable "causes" the other is not particularly clear as this is a "chicken or egg" issue. Until a more objective measure of training adequacy is developed and used in a fashion different from that of this particular investigation, the question of causality will remain unanswered.

Teacher Stress and Central Life Interest

Are teachers who prioritize their work over their personal lives more likely to experience stress than those whose personal lives are more important? This question was investigated by Zacherman (1984) when he surveyed 244 New York City teachers in terms of their central life interests and stress levels. Using the Central Life Interest Questionnaire (note Zacherman, 1984) and the initial form Teacher Stress Inventory, Zacherman established that those special education and regular teachers who prioritized their work life were more likely to report significantly larger TSI strength scores than were those primarily interested in their personal lives. It was hypothesized that removing one's self from life in general also disrupts naturally occurring systems of and structures for support that are not typical of the workplace. Accordingly, stress levels tend to be higher.

Teacher Stress and Supervisory/Peer Support

Are teachers who receive support more likely to experience lower stress levels than those not receiving support? Fimian investigated the receipt of both peer and supervisory support experienced and reported by 1,107 Vermont and Connecticut teachers. The results, reported in greater detail in Fimian (1986b, 1986c), confirmed the often-accepted yet usually untested assumption that the receipt of peer and supervisory support would act as a moderator of teacher stress. Evidently, teachers are more likely to receive aid and support from peers in time of stress than they are to receive it from supervisors: two of every three teachers reported not receiving supervisory support. Peer support, though discussed more, has been researched less. Only a small minority -- 10% of the teachers -- did not receive peer support.

It is also apparent that supervisory support plays a major role in moderating the perceived strength of teacher stress. Those not receiving supervisory support experienced significantly stronger Personal/Professional Stressors, Professional Distress, Discipline and Motivation problems, and Emotional and Physiological-Fatigue Manifestations, as well as the Total Strength of Stress than did those who did receive support. Those not receiving peer support reported significantly stronger Professional Distress in comparison to support recipients, with the nonrecipients experiencing stressful events as being significantly stronger than did the recipients of peer support.

In other work, Courtney (1987) compared 28 teachers not receiving administrative support with 103 teachers receiving such support and found results similar to Fimian's. Except for the initial form TSI subscale Biobehavioral Manifestations, those not receiving administrative support experienced significantly stronger stress related to each of the other subscales and significantly more stress overall than did the supported teachers. Only 21% of this sample reported not receiving such support. Braaten (1987), in his study of role-related stress among 108 special education teachers in Minnesota, found a significant relationship (p = .006) between elevated stress levels and the absence of administrative support; in his study, 57% of the teachers reported receiving administrative support, whereas 43% reported not receiving such support.

Teacher Stress and Job Satisfaction

An analysis was conducted for 539 teachers from the aggregate pool of 3,401 teachers. These teachers were asked to rate their job satisfaction levels as they were completing the TSI. A correlation of .17 (p = .001) between the Total Strength Score and teacher dissatisfaction was established indicating that those teachers most dissatisfied with their jobs were also those subject to the strongest stressful experiences. Using similar analyses and scales, Honaker (1987) found that, of 131 West Virginia teachers, those who reported being less satisfied with their teaching roles were also those experiencing significantly more stress in terms of each of the TSI subscales. Similarly, of Braaten's (1987) sample of 108 special education teachers from Minnesota, those reporting less job satisfaction also reported significantly stronger stress than teachers satisfied with their jobs. Apparently, job dissatisfaction is only somewhat related to teacher stress-teachers who are stressed on the job are not necessarily those most dissatisfied with their jobs.

Future work with well-established satisfaction constructs will need to be conducted to clarify this relationship. Teacher stress and job stress. An additional analysis was conducted for 539 teachers from the aggregate data pool. These teachers were asked to rate their job stress levels as they were completing the TSI. A correlation of .51 (p = .001) with the Total Stress Score indicated that those who rated their jobs as very stressful were the same teachers who reported the strongest on-the-job stress as defined by the TSI. Using similar analyses and scales, Honaker (1987) found that, of 131 West Virginia teachers, those who reported being more highly stressed on the job were also those experiencing significantly more stress across each of the TS1 subscales.

Teacher Stress and Substance Abuse

Is teacher stress at all related to the use of over-the-counter (OTC), prescription (P), and alcoholic (A) drugs? This question was investigated based on the data collected from 1,788 teachers from Vermont, Connecticut, and New York (sample sizes = 365, 371, 371, 435, and 249, respectively). It was determined from these data that the strength of occupational stress plays a role in supporting the use of OTC, P, and A drugs (Fimian, Zacherman, & McHardy, 1985). It was also apparent that (a) the strength of need to use such substances is significantly related to the perceived strength of on-the-job stress -- the stronger the stressful events are perceived to be, the stronger the reported need to make use of substances that would act as stress buffers; (b) 1 in every 20 teachers reported great to major need to use OTC or P drugs to ameliorate stress, whereas 1 in 10 reported a great to major need to use alcohol in the same capacity; (c) 1 in 30 teachers used OTC and/or P drugs on a daily or near-daily basis, whereas 1 in 10 routinely used alcohol during times of stress; and (d) the strength of stress can be used to estimate the degree of need to use stress-buffering substances. In short, OTC, P, and A substances were used by many teachers in the face of stressful events. Also, a small percentage (e.g., 5% to 10%) routinely used and felt the need to use substances in order to reduce stress to more manageable levels. Apparently, the use of drugs is significantly related to the strength of stressful on-the-job events experienced by teachers. Some caution should be exercised when considering these data, however; typically, substance use and abuse tends to be underreported in self-report studies. In the case of this investigation, there was little opportunity to address or control for this issue.

Teacher Stress and Physiological Symptoms

It was hypothesized that the presence of strong levels of teacher stress would be related to the frequency with which psychosomatic symptoms were experienced. Based on the Connecticut teachers' data and reported in an unpublished manuscript (Fimian, 1986e), the relationship between teachers' perceptions of stress and the physiological events they experienced was established. Overall, the teachers reported somewhat to moderately frequent psychosomatic disorders. Frequencies were reported for each of 16 disorders: e.g., stomach acid, cramps, and pain; racing heart; feelings of increased blood pressure; headaches; voice loss; cold sweat; physical exhaustion and weakness; nausea; rapid breath; dizziness; fatigue; back pains; and decreased appetite (Belcastro, 1981; Bloch, 1978; Gmelch, 1977; Knab, 1982); each was significantly related to one another. Teachers who were apt to experience one of the psychosomatic symptoms were very likely to experience at least some of the other 15. Those who frequently experienced one or more of the symptoms were also very likely to experience the rest frequently. When the psychosomatic disorders were inspected in light of reported stress levels, two consistent relationships were noted: (a) teachers under stress experienced significantly more frequent psychosomatic disorders than did those under relatively little stress and (b) teachers who experienced frequent psychosomatic disorders were generally under significantly more stress than those who did not experience symptoms.

Additional work with the correlates of physiological symptoms of stress and rated stress levels has been conducted by Golaszewski et al. (1984). In this investigation, 118 teachers from New York were asked to complete the TSI, and then had their blood pressure taken. In this investigation, significant relationships between teachers' systolic blood pressure and Emotional Manifestations (r = .29, p = .05) and Physiological-Fatigue Manifestations (r = .32, p = .05) were noted. Since only the TSI manifestations subscales were used in this investigation, only future studies will confirm the relationships between biological indicators and the other TS1 stress scores.

Teacher Stress and Anxiety

It stands to reason that the stress experienced by teachers would also be related to their anxiety levels; The more anxiety experienced, the stronger the stress. The TSI has been used in combination with the State-Trait Anxiety Scale (Spielberger, 1973) with three different samples from Georgia (ns = 39, 39, and 10). The Total Stress Score was found to be related to teachers' state anxiety levels: rs = .07 (p = trend),.49 (p = .001), and .87 (p = .05). The Total Stress Score was also found to be related to the teachers' trait anxiety levels (sample n = 10, r = .93, p = .05), indicating that those teachers who are naturally anxious (Le, who score highly on the trait anxiety measure) and who find themselves in stressful situations (e.g., who score highly on the state anxiety measure) are also those who report the strongest stressful experiences associated with teaching.

Teacher Stress and Social Readjustment

Are teachers who experience stressful life events more likely to report stressful work-related events? It was hypothesized that a small yet significant correlation would result between work stress and life stress. The TSI and the Social Readjustment Scale were used to survey 23 of the 266 North Carolina teachers listed in Table 10 . The Social Readjustment Rating Scale (Holmes & Rahe, 1967), a commonly used measure of the degree to which stressful life experiences occur to adults, was significantly related to Discipline and Motivation (r = .46, p = .05), Emotional Manifestations (r = .42, p = .05), and the Total Stress Score (r = .42, p = .05) of the TSI. A multiple R of .18 (p = .05) indicated that, after accounting for variance attributable to Type A/B personalities, locus of control, and need deficiencies, stressful life events still accounted for a significant percent of variance related to the strength of teacher stress. From these data it is apparent that work-related and life stress are related to one another to at least a limited degree. Work with larger samples, though, is needed to verify and clarify such relationships.

Teacher Stress and Tedium

Is there a relationship between teacher stress levels and the tedium experienced during their day-to-day routines? First investigated by Pines, Aronson, and Kafry (1981), tedium was found to be related significantly to the burnout model developed by Maslach and Jackson (1981). It stands to reason that classroom tedium would also be significantly related to teacher stress. The Tedium Measure was used with a subsample (n = 21) drawn from the 266 North Carolina teachers. The teachers completed both the TSI and the Tedium Measure; from these data, it was determined that the Total Stress Score of the TSI was significantly related to the Tedium measure subscales Physical Exhaustion (r = .77, p = .001), Emotional Exhaustion (r = .80, p = .001), and Mental Exhaustion (r = .78, p = .001), as well as to the Total Tedium Score (r = .83, p = .001).

Teacher Stress and Principal Management Style

Do the ways in which principals express leadership, use power and authority, arrive at decisions, and interact with teachers (Sergiovanni & Elliot, 1975) have a bearing on teacher stress levels? Using the Managerial Style Questionnaire developed by McBer and Company (1980), Courtney (1987) investigated these issues by examining the relationships between leadership styles and teacher stress as measured by the TSI and reported by 267 North Carolina teachers. Hypothesizing that some styles of management would produce more stress for teachers whereas others would result in less, Courtney determined that principal management style was indeed significantly related to teacher stress at the .001 level -- the management coaching style was the least stressful to teachers, and the pacesetting style the most. Courtney also noted that principals with coaching styles tended to display a concern for high performance standards and saw their job as helping or showing teachers how to improve their performance and encouraging professional development and shared decision making. Accordingly, they directed by asking their teachers to set their own goals.

Pacesetters, on the other hand, tended to set high standards and to lead by example, have trouble delegating and taking responsibility, become coercive when teachers experienced difficulties, have little sympathy for poor performance, and not develop or support their teachers. Given the work climate generated under such a management style, it is clear why teachers under the coaching principals would experience and report less stress than those under pacesetters.

Teacher Stress and Principal Leadership

Is there a relationship between the ways in which principals lead their staff and teacher stress levels? Courtney investigated this relationship by surveying 267 teachers with respect to their stress levels and their principals' leadership style, as defined by the Principal Management Instructional Rating Scale (note Courtney, 1987). This scale measures 11 job functions that reflect the areas of principal responsibility in his or her role of instructional leader. Instructional leadership of the principal was found to be significantly related to teacher stress levels, particularly with respect to promoting a positive learning climate, developing and enforcing standards, providing incentives for teachers, and promoting professional development. The absence of these factors was significantly related to elevated teacher stress levels.

Teacher Stress and Stress Inoculation Programs

Some limited data exist that indicate that teacher stress, as defined by the TSI, can be modified through the use of stress inoculation training programs. Cecil (1987) and his advisor Susan Forman used the TSI to measure the differential effects that stress interventions would have upon teachers. Fifty-four South Carolina teachers were randomly assigned to one of three experimental conditions -- stress inoculation training, coworker support group, and a no-treatment control group. Teachers were assessed before treatment, immediately after treatment, and 4 weeks later. Programs were conducted for each of the treatment conditions for 90 minutes per week over a period of 6 consecutive weeks. The stress inoculation group was effective in reducing stress related to Professional Distress, Personal-Professional Stressors, Discipline and Motivation, and Emotional Manifestations, whereas the coworker support group was not.