The day-to-day reality of teacher turnover in preschool classrooms: an analysis of classroom context and teacher, director, and parent perspectives.

The day-to-day reality of teacher turnover in preschool classrooms: an analysis of classroom context and teacher, director, and parent perspectives. The purpose of the current study is to examine teacher turnovercomprehensively by triangulating the experiences of teachers, directors,parents, and children through actual, "real-time" turnovertransitions. We intentionally examined turnover with a small sample size(N = 13 classrooms) to facilitate comprehensive data collectionutilizing multiple qualitative and quantitative measures and to gain arich understanding of the implications of teacher turnover on classroomquality, staff, and the parents and children with whom they work. Thestudy utilized an interpretive paradigm to illuminate and juxtapose theexperiences of teachers, directors, parents, and children through theturnover transition--as teachers departed and their replacements began.Strategies used to respond to turnover were identified at the programand classroom levels. Implications for early childhood classroom qualityand policy are discussed. Keywords: early childhood education, child care, teacher turnover,child care quality ********** The high rate of teacher turnover continues to be a concern for thechild care industry. Research reveals that an estimated 82% of childcare teachers employed in 1994 and 76% employed in 1996 were no longerretained in the field by the year 2000. Scholars characterize theprofession as "alarmingly unstable" (Whitebook, Sakai, Gerber,& Howes, 2001), with this instability affecting children, theirfamilies, and, of course, child care teachers. The purpose of the current study was to examine the implications ofteacher turnover through a mixed-methods approach, one that includedmultiple perspectives from within the classroom context. Classroomobservations, interviews, teacher and director survey self-reports, andresearcher fieldnotes served to facilitate an understanding of thedynamic nature of teacher turnover and its implications for directors,teachers, children, and parents. A comprehensive analysis of teacherturnover was conducted by juxtaposing the experiences of teachers,directors, parents, and children through actual, real-time turnovertransitions--as teachers left and their replacements began.Comprehensive data collection was facilitated by intentionallycollecting data from a small sample (N = 13 classrooms). Multiplequalitative and quantitative measures were implemented to gain a richunderstanding of the implications of teacher turnover on classroomquality, staff, and the parents and children with whom they work. The study utilized an interpretive paradigm to illuminate andjuxtapose the experiences of teachers, directors, parents, and childrenthrough the turnover transition. The ontological assumption of theinterpretive paradigm is that "reality is sociallyconstructed" (Mertens, 1998, p. 11). From this perspective,understanding the perceptions and experiences of teachers, directors,parents, and children--as they interrelate in ways that support andresist each other--during turnover is critical to addressing importantimplications for classroom environments. By collecting data during realtime (rather than from prospective or retrospective accounts) andthrough the use of qualitative and quantitative measures, a richunderstanding of how programs and parties (directors, teachers, parents,and children) react and adapt in ways that support and resist each otheris fostered. Classroom quality as well as teaching and workingrelationships among early childhood staff and among the parents andchildren with whom they work are examined. By compounding data sourcesand triangulating perspectives (i.e., teachers, directors, parents), wewere able to learn about the multiple realities of turnover, as well asways that may serve to help teachers, directors, parents, and childrenthrough such transitions. BACKGROUND Research has established a link between teacher turnover, globalquality, and child outcomes. Centers with higher rates of teacherturnover show lower levels of global quality and less appropriateteacher-child interactions (Phillips, Mekos, Scarr, McCartney, &Abbott-Shim, 2000). Furthermore, in programs with high turnover,children have lower levels of language, cognitive, and socialdevelopment (Howes & Hamilton, 1993; Howes, Hamilton, &Philipsen, 1998; Howes, Phillips, & Whitebook, 1992; Howes &Smith, 1995). For example, Whitebook et al. (2001) conducted alongitudinal study of staffing and quality. The examination included 75centers with data collected at three points in time (1994, 1996, and2000). Average turnover per year was 30%, with a range of 0% to 100%.The study further revealed no difference in turnover rates betweencenters accredited by the National Association for the Education ofYoung Children (NAEYC) and those that were not accredited. More stablecaregivers were associated with a more secure relationship with thechildren in their classroom. Low turnover, combined with well-educatedstaff with specialized training in child development, was related tohigher classroom quality. However, the link between teacher turnover and child outcomes isquite complex. Centers with higher teacher turnover rates and lowerlevels of child outcomes also have higher child-to-adult ratios, fewereducated teachers, and are generally characterized as poor-qualityprograms (DeVita, Twombly, & Montilla, 2002), which makes itdifficult to attribute poor child outcomes strictly to teacher turnover.There has also been extensive research about the reasons for teacherturnover (Hale-Jinks, Knopf, & Kemple, 2006), with a particularfocus on low salaries, lack of benefits, and difficult workingconditions, resulting in low morale, stress, and job burnout (Curbow,Spratt, Unagretti, McDonnell, & Breckler, 2001), as well ascompromised organizational climate (Bloom, 1997). Teacher Turnover and Work Environments Theorists in organizational behavior science suggest three factorsthat influence employee retention and turnover (Whitebook & Bellm,1999). Employees are likely to remain in their current positions whenthey have adequate pay and benefits, good work environments withprofessional development opportunities and satisfactory relationshipswith co-workers, and a good match between their work responsibilitiesand the job expectation and training. Low teacher salaries andunsatisfactory benefits in child care have long been associated withhigh rates of turnover (Goodman, Brady, & Desch, 1987; Phillips etal., 2000; Stremmel, 1991; Whitebook et al., 2001). Goodman et al.(1987) reported that salary increases in Head Start programs resulted ina sharp decrease in teacher turnover, from 65% to 35%. Olenick (1986)also found a similar pattern of better compensation and higher retentionof teachers in child care programs. More recent research (Whitebook& Sakai, 2003) examined the relationship between teachers who leaveand those who stay. The study revealed that directors were most likelyto leave because of low wages, whereas teachers, especially highlyqualified teachers, were more likely to leave not only because of lowcompensation but also because they could not find a compatible level ofeducation among coworkers in their work environment. Data from the Center for the Child Care Workforce (CCW; 2004)underscore the relationship between inadequate compensation for teachersand turnover rates: The average hourly rate of pay for a child careteacher in 2006 was $9.05 per hour, with turnover ranging from 25% to40% (CCW, 2004). Not surprisingly, this rate of pay was higher than only18 other occupations and was less than what service station and lockerroom attendants earn (Bureau of Labor Statistics, 2007). Child careteachers earn slightly more than one-third of the salary of a publicschool kindergarten teacher, even though many have the same degree andperform somewhat similar teaching activities. The problem of teachersalaries is so intractable that many in the field, including directservice providers, have become somewhat complacent about turnover. Theyhave come to believe there is nothing that can be done about salariesand, thus, turnover, so they regard the high turnover rates asinevitable. Furthermore, Manlove and Guzell (1997) examined job turnoverin 169 child care workers in rural and semirural communities inPennsylvania. They cited high turnover as one of the greatest barriersto providing high-quality child care. Turnover in their study averaged23% over a 12-month period. However, 41% of the teachers in theseprograms indicated that they expected to leave their jobs in the nearfuture. Reasons for leaving their jobs included advancement (changingjobs) and returning to school. None of the participants in the studyleft the programs to accept employment in other early childhoodprograms. The authors also found those with less experience and thosewith other job options were most likely to leave their positions. Teacher Turnover and Child Outcomes Turnover also can have a serious impact on teacher-childrelationships within child care classrooms. Howes and Hamilton (1993)examined the relationship between caregiver-child attachment and teachersensitivity. They found that children who were classified as securelyattached experienced more teacher sensitivity and involvement, whereaschildren classified as ambivalently attached experienced less teachersensitivity and involvement. Children in the avoidant categoryexperienced the least sensitivity and involvement. These findings arerelevant as we consider the relationships that are severed betweenteachers and children when teachers leave their position. Furthermore,there is evidence that the development of positive relationships canaffect children's long-term academic and social development. Hamreand Pianta (2001) found that children who had difficulty formingpositive relationships with kindergarten teachers also had lateracademic and behavioral problems in school, especially those inteacher-child relationships marked by conflict and overdependence. Overall, these findings from previous research demonstrate thatturnover can affect classroom quality and child outcomes, due to changesin relationships within the classroom. However, most of the studiesconducted on turnover have been retrospective. Thus, these studies haveexamined program quality with the rate of turnover in the past year,wages and benefits, and working conditions. The current studycontributes to the field's understanding of turnover as it (1)examines the actual transition in the turnover process as one teacherleaves and another replaces her, (2) exemplifies the complexity ofteacher turnover and its effects on classroom quality, and (3)demonstrates how turnover influences the lives of directors, teachers,parents, and children. Whitebook and Sakai (2003) identified three typesof turnover: (1)job turnover, in which a teacher leaves a child carefacility; (2) position turnover, in which a teacher moves to anotherclassroom within the center agency; and (3) occupational turnover, inwhich a teacher leaves the child care field. In the current study, allthree types of turnover were included, as was temporary turnover, suchas extended maternity or sick leaves. RESEARCH QUESTIONS To understand the impact of turnover on the quality of classrooms,several questions guided our inquiry: (1) What are the day-to-dayexperiences of teachers, directors, and parents during turnovertransitions? (2) How is classroom quality compromised as a result ofturnover? (3) What areas of the classroom are most affected (e.g.,materials/activities, language/interactions, global quality) byturnover? and (4) How are relationships among children, parents, andstaff affected by turnover? METHOD Procedure Understanding quality ratings of the child care centers. In anattempt to increase and maintain the quality of child care centers, thestate of North Carolina has implemented a star rating system. Centerscan obtain a rating from one to five stars, based on two essentialcomponents: program and educational standards. For program standards,the centers are assessed on their operating and personnel policies,activity areas in the classroom, square footage area per classroom,staff-child ratios, and their ratings on standardized instruments, suchas the Infant/Toddler Environment Rating Scale-Revised (ITERS-R; Harms,Cryer, & Clifford, 2006) and Early Childhood Environment RatingScale-Revised (ECERS-R; Harms, Clifford, & Cryer, 1998). Theeducational standards considered for licensing include the education andexperience of teaching and administrative staff and the number ofteachers with child care credentials (North Carolina Department ofHealth and Human Services, 2010). The greater the number of stars, thehigher the quality of the program. In the current study, recruitingcenters with different star ratings was important, because programs ofvarying quality may handle turnover in different ways, affectingteachers, directors, parents, and children differentially. Recruitment of centers. Center directors from child care centersranging in quality on North Carolina's five-star rated license werecontacted and asked to participate in the study. A purposive samplingtechnique was employed to recruit centers. To begin, a list of childcare centers in the county of the study was gathered from the NorthCarolina Division of Child Development website. A simple search on thewebsite using the county names led to a list of child care centers withtheir star rating. Initially, centers were randomly contacted from thatlist. Directors who consented or demonstrated willingness to be involvedin the study were then short-listed. Directors of child care centers onthe shortlist were provided with a full explanation of the study. Ifthey were interested in the concept of the study or anticipated apossibility of turnover in their center, they were asked to contact theresearcher or give their verbal consent to be contacted twice a month tocheck on the center's turnover status. Once a situation of turnoverwas identified, the researcher visited the child care center, providingfurther explanation about the study, answering questions, and providingconsent forms to the director and all teaching staff in the classroomexperiencing turnover, including the departing and any remainingteachers in the classroom. One week after the new (or replacement)teacher started in a classroom, the researcher made another visit to thecenter, providing her with an explanation of the study, answeringquestions, and requesting consent to continue the study while she was inthe classroom. Over a 2-year period, 38 centers were contacted, 11centers agreed to participate, and 9 of those actually experiencedturnover. The star rating of the nine centers that experienced turnoverranged from three to five stars. Participant characteristics. Overall, 34 teachers participated inthe current study (see Tables 1 and 2 for teachers' demographics).Most of the teachers were European American, followed by AfricanAmerican and other ethnic groups. Most of the teachers had some collegeeducation. Teachers were employed within the child care field for anaverage of 78.39 months and earned an average hourly wage of $8.82. Mostof the teachers received full benefits from their employee, with ahandful receiving complete retirement packages. Similar to the teachers,most of the directors were European American; there were two AfricanAmerican directors and one American Indian director. Only one directorhad a graduate degree, whereas the other directors had either a 2- or4-year degree (see Table 3 for directors' demographics). Study implementation during turnover. Directors sometimes were ableto provide or predict future turnover as a result of forewarneddepartures, such as maternity leaves. At other times, we had to respondmore quickly as teachers gave relatively short notices. Once theresearch team was made aware of a turnover situation, consent forms wereprovided to the director and teaching staff of the classroom whereturnover was taking place. Upon consent, during the last week ofemployment, teachers leaving a program were identified as"departing teachers." Departing teachers were interviewedduring their last week of employment, their classrooms were observed,and they completed survey measures. New teachers were considered thoseteachers replacing the departing teacher and may or may not have beennew to the center. That is, in some cases, departing teachers werereplaced by a "floater" or another teacher working in thecenter. In other situations, new teachers to the program were hired asreplacements. Approximately one week after the new teacher replaced thedeparting teacher, she was observed and interviewed by an independentresearcher and asked to complete the survey measures. Thus, new teacherswere given a week in their new position before we began collecting data.Giving a week to the new teachers in their own classroom ensured thatthey could complete the various measures implemented in the study in anappropriate manner. That is, the work environment surveys required thatthe teachers have some knowledge about the center they are working for.The remaining teachers in the classroom (e.g., coteacher, assistantteacher) were identified as the teachers who were consistent through theturnover transition. They worked with the departing teacher and the newteacher with the same group of children. These teachers also wereinterviewed and completed the survey measures once during the datacollection process. A sample of 13 classrooms, including 34 teacherswithin nine centers, resulted. Two of the teachers who participated inthe project were initially remaining teachers, and, later in theproject, left the programs and were counted as departing teachers. Atthese different times, these two teachers participated fully in theproject according to their teacher status. Teacher demographic data andemployment information are reported in Tables 1 and 2, respectively,with director demographic data reported in Table 3. Parent recruitment. At the time the study began in each of theclassrooms, information about the study and consent forms were sent homewith children requesting consent for parent interviews. Interestedparents were interviewed at the center or by telephone after the newteachers' arrival. Almost one half of the parents (42.3%) from eachclassroom consented to participate in the interviews. Of these parents,65.4% were interviewed. Qualitative Measures Interviews. Semistructured interviews (CCW, 2000) were conductedwith directors, teachers, and parents. Trained interviewers asked open-and closed-ended questions to elicit in-depth responses from theparticipants (see the appendix). Overall, each interview lastedapproximately 30 to 45 minutes. Fieldnotes. The independent observers (from the observation withthe departing teacher and the observation with the new teacher) keptfieldnotes about the procedures that participating centers implementedto maintain consistency in the classroom during the teacher turnovertransition. Through a constant comparative method (Glaser & Strauss,1967; Rennie, 2006), the fieldnotes assisted researchers in recognizingdistinctions in the way centers dealt with and responded to turnover. Aconstant comparative method is a qualitative approach in grounded theoryto code data while simultaneously creating definitions through analysis. Quantitative Measures Early Childhood Environment Rating Scale-Revised. The ECERS-R(Harms et al., 1998) is a widely used measure of global quality.Additionally, two factors have been found to differentiate the scaleinto two dimensions of quality: Activities/Materials andLanguage/Interactions (Cassidy, Hestenes, Hegde, Hestenes, & Mims,2005). The Activities/Materials and the Language/Interactions factorsrepresent subsets of the total scale and are utilized in the currentstudy. That is, the total ECERS-R score, subscales, and two factorscores were used to examine changes in quality from pre- topost-turnover. Past research has excluded the Parents and Staff subscaleto focus this instrument as a measure of "child-related items"(de Kruif, McWilliam, Ridely, & Wakely, 2000, p. 254). Because thecurrent study included more comprehensive measures to capture theperspectives of parents and staff, we felt comfortable excluding theParents and Staff subscale from the ECERS-R. This step is consistentwith numerous studies of quality using the ECERS-R. All of thereliabilities for internal consistency were high; Cronbach's alphafor the overall scale was .95. The Activities/Materials factor alpha was.92, and the alpha for the Language/Interaction factor was .93. Due tothe small number of participating centers, the ECERS-R data were usedfor descriptive purposes only. Student-Teacher Relationship Scale. The Student-TeacherRelationship Scale (STRS; Pianta, 2001) is a 28-item self-reportinstrument that examines teachers' perceptions of theirrelationships with the children they work with and children'sinteractive behavior toward them. The scale examines three factors:Conflict, Closeness, and Dependency. Some examples of items are "Ishare an affectionate, warm relationship with this child" and"This child is overly dependent on me." Items are scored on aLikert-type scale ranging from 1 (definitely does not apply) to 5(definitely applies). Cronbach's alpha for this measure was .75.This instrument was included for all teachers, including new teachers,to demonstrate the possibly notable difference in relationship betweenthe teacher who departed and the newly arrived teacher. Although thescale recommends a short period of familiarity between teacher andchild, the purpose of the study was to examine the loss for childrenwhen a teacher leaves. Data Analysis As a mixed-methods study, qualitative data were analyzed using aninterpretive approach, whereas descriptive and inferential statisticswere used to analyze quantitative data. Qualitative data analysis. An interpretive approach was used tocode the qualitative data (Creswell, 2005). Interviews were transcribedand read to get a "general sense" (Creswell, p. 231) of thedata by three independent coders. In reviewing the transcripts, allthree coders wrote memos that informed an a priori coding scheme, whichwas identified collectively through discussion among the three coders.The coders then independently applied the coding scheme to thetranscripts. For example, text was highlighted and labeled with anappropriate code. Any new codes that were identified were recorded asmemos for future analysis. A continuous process of reading and assigningcodes to the transcripts among the three coders ensured that themes wereapplied appropriately. Discrepancies were resolved through discussion.For example, in the first round of coding, the coders came up withseveral descriptions for children's behavior as a result ofturnover; these included "children's misbehavior,""testing their boundaries," "acting out,""sense of sadness," and "frustration/anxiety." Afterdiscussion regarding the details of the statements within thetranscripts, as well as comparisons across transcripts, these codes werecondensed to an overall theme of "children's behavior"and subthemes of "sense of sadness," "acting out,"and "where is so and so." The results from these analyses provide information regarding theday-to-day experiences of turnover from the perspectives of thedirectors, teachers, and parents involved in the turnover transition.Additionally, these results indicate changes in relationships amongteachers, parents, and children associated with turnover. Individual andshared perspectives of these experiences and changes in relationshipsare discussed in detail in the qualitative results section. Quantitative data analysis. Due to the small sample size, it wasonly appropriate to identify mean and standard deviations for ECERS-Raverage, subscale, and factor scores. Comparisons among the descriptivestatistics were drawn across proactive and reactive centers andindividual classrooms, pre- and post-turnover. MANOVA was employed tocompare differences between teachers (departing, remaining, and new)scoring on the various STRS factors, because the sample size allowed forsuch comparisons. For these analyses only, follow-up post hoc analysisrevealed which group of teachers differed significantly from each other. RESULTS Distinctions Between Proactive and Reactive Centers Based on the fieldnotes and memos from the interview transcripts,distinctions among centers were identified from the policies andprocedures implemented during the turnover transition. In response tothese distinctions, centers were categorized as either"proactive" or "reactive." Utilizing the constantcomparative method, these ideas were solidified as important qualitiesto consider in the way that child care programs navigated thecomplexities of turnover, and subsequently were important to theinterpretation of the results. Specifically, child care centers that were identified as proactivehad systems in place to minimize the disruption of turnover. Forexample, proactive centers were more likely to have a"floater" position established in the program, someone withwhom children and parents were familiar and who was accustomed to thecenter, its policies, daily routines, and instructional environment.Often, this "floater" was placed in the classroom before thedeparting teacher left the program and remained in the classroom untilafter a transition period with the new teacher. Conversely, child carecenters that were more reactive responded to turnover by shiftingcurrent staff or children to maintain state-required ratios or relied onthe use of multiple substitutes in the classroom. For example, onecenter that was classified as reactive responded to turnover bycombining two classrooms after the teacher left, shifting children andteachers to maintain state ratio requirements. This reactive approachnot only resulted in children and families losing their teacher, butalso meant some of the children experienced an unfamiliar classroom inaddition to multiple substitutes. In the current study, four centers(five classrooms) were identified as proactive and five centers (eightclassrooms) were identified as reactive. Although we divided child careprograms that were more reactive or proactive to turnover into twogroups, we acknowledge that this concept may fit best on a continuum,with centers having varying degrees of reactivity and proactivity.Additionally, some of the qualitative results were similar acrossproactive and reactive centers. However, the distinctions betweenproactive and reactive centers are indicated where appropriate. Qualitative Results Teachers' perceptions of turnover Instructional changes: "You cannot teach and you cannot helpas many children." Teachers discussed the impact of turnover ontheir own work and in relationship to the experiences they were able toprovide to children. The departing, new, and remaining teachers acrossreactive and proactive centers all talked about challenges they faced intheir work environment as a result of turnover. For example, a departingteacher stated, "A lot of times, working with someone, you getclose to 'em.... It's kind of hard when they do leave,'cause you learned a lot about 'em." Another departingteacher recognized the implications of teaching with fewer teachers inthe classroom. She stated, "It's just when you don't haveenough people in this room, you cannot teach and you cannot help as manychildren as you would like." The teachers remaining in their classrooms during the turnovertransition felt largely depended upon and reported increased stress. Forexample, one teacher stated, "After she left and I had all thoseother teachers to come in, it just wasn't the same; it was likeeverything was put on me." Another remaining teacher noted, "Iwouldn't have help, you know, for some time." Additionally,teachers described an increased workload. For example, one teacherstated, "It makes it more difficult to just work with a new person;they don't know the routine." Further, it impacted "myability to plan, because my time is taken with other things in theclassroom instead of being able to take the time to plan our classroomactivities." The new teachers also talked about an increased workload as aresult of turnover. For example, a new teacher noted, "You have topick up the slack for whoever is gone." Interestingly, the newteachers also noted the increased responsibility of remaining teachersto orient the new teacher (in this case, them) to the position. One newteacher explained how a remaining teacher has to "basically do thesame thing that [she] did with the previous person." Accordingly,another new teacher reflected on learning the logistics and routines ofthe classroom: "When the [new teacher] come[s], you have to showher the routine, you have to show her the schedule and stuff." Nomatter the status of the teachers or reactivity of the program, turnoverresulted in an increased workload, making it more difficult to teacheffectively. Classroom management: "Just trying to regulate the wholeclassroom." Teachers also discussed the complexity of keeping dailyroutines consistent, maintaining a positive climate through classroommanagement, and having to compromise instructional activities in theface of teacher turnover. However, the difficulties associated withclassroom management were expressed primarily by teachers in reactivecenters. One teacher said, "It disrupts your routine andeverything.... If you had something planned you know and somebody staysout, you can't do what you planned on doing." Furthermore, ateacher remarked, "It's harder, especially when you havesomebody who comes in who might not know the routines of theclassroom.... They need to get used to the classroom, used to thekids." Similarly, another teacher stated, "Everything prettymuch falls on that one person because.., if they do send someone in,they don't know the routine." As another teacher described it,"You have to again kind of train the person." Similarly,"just having to tell a sub what they need to do" was an addedstress to the day. Moreover, classroom management was described as especiallydifficult during turnover, affecting the activities and the interactionsteachers were able to have with children. For example, as one teacherstated, "Just being able to teach the children is hard, becauseit's hard for me to get involved with a small group of childrenwhen I have to maintain the classroom as a whole.... I've had to domore managing of the classroom." Another teacher noted, "Ican't teach them or sit with them like I want to because I have toshow the other teacher, you know, how to go move around the classroomand what to do." For one teacher, instructional activities were nota possibility because so much time was spent on daily routines. As shesaid, "You can't do the activities ... when you're all byyourself in this room.... You have to change them, feed them, nap them,and that's it." Unfortunately, as a result of turnover,teachers were less able to regulate their classroom and create apositive learning environment for the children. Teachers' perceptions of the impact of turnover onchildren's behavior In addition to trying to maintain a sense of normalcy in theclassrooms during turnover, teachers also are meeting the varyingdemands of children's reactions to the change incurred by turnover.Departing, remaining, and new teachers in proactive and reactive centersshared some similar ideas about how turnover affects children. Forexample, "behavioral changes" in the children emerged as atheme from all teacher interviews, regardless of their role in thetransition. Additionally, sub-themes captured children questioning"where is [the teacher]?," having a "sense ofsadness," and "acting out" as reactions to turnover. Confusion: "Where is so and so?" Across proactive andreactive centers, departing teachers reported that turnover was"confus[ing] at first" to children, as "they don'tunderstand" and "they are looking for that teacher." Oneteacher stated, "They wonder where the teacher's gone and areand where are they going to, are they coming back." Another teacherdescribed the experience as being "very confusing for them, theydon't understand why, suddenly their teacher was here and then shewasn't.... I think the first few days, they may be asking,'Where is so and so?'" Similarly, a teacher noted,"They get curious and ask a lot of questions, 'Well, where isshe?,' 'When is she coming back?'" Remaining and new teachers expressed similar ideas related tochildren being confused and inquiring about the teacher who left. Forexample, a remaining teacher noted how "It takes time to get usedto somebody new ... they're still kind of like, 'Is she comingback?'" Another remaining teacher described children as"needing consistency and when there is not, the children seem to be... confused: 'Who's the teacher?' 'Who do I listento?' "Another remaining teacher reflected, "You become apart of their life, you know, when they come see you every day, then allof a sudden you just disappear: 'Where did she go?,''What happened?' Kids stress just like adults stress."For one new teacher, the children "just want to know why" aparticular teacher is leaving. As another new teacher observed,"They are very confused and they want to know where that teacher isand what happened." In general, no matter the reason forteachers' departures, teachers reported that children seemed to beconfused about why teachers were leaving and concerned about them notreturning. However, in proactive centers, the children were givenpositive strategies, such as writing letters to the departing teacher,to help them with the transition. Sorrow: "Sense of sadness." Regardless of centerclassification, the departing and the remaining teachers also observedchildren experiencing "sadness" from the loss of a teacher.One departing teacher noted, "I think they're genuinely sadthat she's not there." Another departing teacher said, "Ithink the children are very sad.., they're more emotional.They'll come up to you and they'll tell you, 'I'mgoing to miss you' and they give you things, 'I made this foryou.'" Teachers also made distinctions among children,indicating that some children will show their "sadness" morethan others. Specifically, one departing teacher described how one childwould "be most hurt by it" because of a close relationshipwith that teacher. A remaining teacher also noted, "They do have asense of sadness.... Sometimes they feel like they've been leftbehind." Another teacher reflected that it was "hard for me asa teacher to get used to it, so I know for the children it'shard." Although a couple of new teachers expressed they saw sadnessamong the children, this was not a prevalent theme; this may be afunction of new teachers wanting children to be happy to have them as ateacher. Frustration: "Acting out." All the teachers describedchildren "acting out," "pushing the limit," and"testing boundaries" during the turnover transition. Forexample, one departing teacher stated, "Children tend to test theirboundaries [with] the new person.... [They are] probably going to have afew more arguments, get much louder, not follow the rules like theywould when their regular teacher is in the classroom, kind of act likethey don't really know what they're supposed to do."Another departing teacher commented, "They'll test whoever isin there.... It's hard for them to regain control.., because thatstable person isn't there anymore, so they are going to startacting out and testing boundaries." Many of the new teachers alsodescribed children "get[ing] distracted, push[ing] your buttons,see[ing] your limits." For example, a new teacher noted, "Ithink they just test whoever is in there; see if they're going tostick with the same rules, if they can get away with something."Overall, teachers identified several ways in which children manifestedin confusion, sadness, or frustration after losing a teacher. Distinctions by role of teacher Although all teachers shared some similar insights, they also wereaffected differentially by the position they played in the transition.For example, the departing teachers emphasized that if children knew theincoming teacher, it "wasn't as big of a deal" and that,over time, children "kind of adapt," suggesting the temporarynature of the impact. This idea of a temporary impact was emphasizedmore by departing teachers than by new or remaining teachers.Understanding the effects of turnover as temporary may have eased theirconcern for the children upon their departure. Additionally, newteachers uniquely expressed the complexities of being new and"learning all the people." One new teacher noted,"I'm trying to get all this in my head so that I don'tmess up." Similarly, remaining teachers talked about changes in theoverall environment and how they took a more active role. For example,one teacher stated, "You feel like that you have to be overeverything." These experiences of teachers echo the idea that allteachers are affected by turnover; however, the nature of this impactmay differ according to the role in the system. Directors' perceptions of turnover Center-wide effect: "Domino effect." Directors describedthe impact of turnover as a "domino effect ... because you'lllose somebody and then there may be a whole shifting of people."Directors indicated that turnover required a reorganization of staffthrough adjusting, shifting, and accommodating. For example, a directornoted, "Everybody sort of has to adjust and shift andaccommodate." Additionally, a director compared turnover to "apuzzle; it has to be put together and it all has to fit and work withthe classroom, the whole center, as well as meet the requirements of thewhole state." All center directors, regardless of the use ofproactive or reactive strategies during turnover, suggested thatturnover created change in the center as a whole. However, reactivecenters held more negative attitudes about the change. A director from aproactive center noted that "a lot of times it [change] can begood." Conversely, one director from a reactive center reflectedthat such change was "not really great for morale." Similarly,a director noted, "It just affects everybody and their wholecohesion." All of the directors mentioned the use of substitutesand floaters. Primarily, directors seemed to rely on their currentteachers and floaters to fill in the gaps in order to maintainstate-required ratios. However, proactive centers had floater positionsthat helped to maintain consistency in classrooms during turnover:"We usually rely on floaters and part-time people that can workmore hours than they're scheduled for." Another director froma proactive center noted, "I feel really fortunate that Ihaven't had to scramble and find somebody that I don'tknow." Conversely, reactive centers were more likely to shiftteachers from other classrooms or move children into other classrooms asa temporary solution. Consistently across proactive and reactivecenters, directors reported, "It is always difficult to find goodsubstitutes." In fact, a director from a reactive center describeddesperately calling on "ladies from my church to come helpme." All directors recognized that remaining teachers must contendwith an extra workload as a result of turnover: "It overworks therest of the teachers because they have an extra burden." No matterthe reason for the turnover or the level of proactive or reactiveresponses, all of the directors noted that changes were felt by everyoneat the centers. Work overload: "It's just too many hats." Directorsalso talked about accumulating extra work during turnover. One directorexplained, "It's just too many hats, you have to wear too manyhats." Another director noted "I have 2-3 people's jobsthen" Some directors discussed working in classrooms in order tocover ratios, taking them away from their administrative work. Forexample, one director stated, "I usually have to cover theclassrooms, so that takes away my time in the office to get thingsdone." Another director noted, "Go[ing] in the classroom meansI have to stop here and then whenever I get a chance, or stay latewhenever everybody's gone, to get it [administrative work]done." Furthermore, directors also described the time it took to find andinterview new hires. For example, one director stated, "Ican't necessarily focus on what I need to focus on during the daybecause ... people walk in the door to fill out an application and itdisrupts what you're in the middle of doing." Another directordescribed "spend[ing] all of my time interviewing people, callingtheir references, making decisions, making staff changes, schedulechanges." Further, a director recognized that "it takes thetime to replace them [departing teachers], it takes time to orient a newperson to their job.... You sort of have to redo things you'vealready done." Additionally, directors from both reactive andproactive centers talked about how the extra work "affect[ed] theirhome life as well as [their] job." Specifically, one directordescribed how she "wind[s] up working sometimes all day on Saturdayor part of Sunday." According to these statements, as directorscope with the domino effects of turnover in their centers, they alsohave to take on more and varied responsibilities, making their jobs evenmore difficult. These increases in workload occurred for directors inboth proactive and reactive centers. Prevention of turnover: "Increase their salary, give thembetter benefits, and provide an assistant teacher." Directorsunanimously talked about needing to "increase staff salary andimprove staff benefits" in order to reduce the rate of turnover.One director described the importance of "supplement[ing] teachersbased on levels of education, experience, and responsibility so thatteachers really are paid what they are worth." Interestingly,directors also emphasized the need to reduce teacher-child ratios.Consistently, directors discussed the complexities of trying to providehigh-quality services to children in the context of the current stateratio regulation. For example, one director working for a corporatechild care program stated, "Companies don't give us enoughhours [for teachers to work] to put two [teachers] in there, because thestate is only 1 to 10 [children] so, without a doubt, it's lower,it's gotta be lower [ratios]." Another director stated,"I would love to have 1 to 3 ratio in infants, but it's l to4.... I'd love to have lower class sizes and group sizes, butthat's the regulations right now and that's what's seenas, you know, good care for children." In accordance, anotherdirector described how "you don't have time to do activitieswith them when you have 18 two-year-olds and there are only two ofyou.... I think we would see a lot less turnover if we had smallerratios and, of course, higher pay." Similarly, one director wishedfor "lower class sizes and staff-to-child ratios because, ofcourse, teachers are able to accomplish more if they are working withfewer children and are able to have the time to devote to implementingthe curriculum as they planned." Additional staff also makes for abetter work environment for teachers, as described by one director:"It's not as difficult for a person to feel like they can takea vacation cause they know there's ... a competent person to coverfor them." Each of the directors noted that money, benefits, andlower ratios would help to prevent turnover; however, they did not haveaccess to the resources that would allow them to implement suchpreventions. Parents' perceptions of turnover Similar to teachers' perceptions, many parents expressed thatchanges in the classroom staff, whether temporary or permanent, affectedthe parent-teacher relationship, as well as the teacher-childrelationship. Specifically, parents in proactive and reactive centersstated that a change in or loss of relationship with teachers resultedin heightened concern for their children's welfare and diminishedquality of communication with classroom staff. With regard to theteacher-child relationship, parents mentioned changes in theirchildren's behavior, emotionally and physically. Parent-teacher relationship: "I had a real rapport andrelationship." Several parents noted that children are not the onlyindividuals who form relationships with teachers; parents also establishrelationships with teachers over time. As one parent stated,"Having the same teacher for a longer period of time gives acomfort level because you establish a rapport. You know that teacher hasa certain amount of history and background with your child." Whenthis relationship ends because of teacher turnover, parents mayexperience difficulty in establishing that trust and confidence in thenew teacher: "This is the person that you have trusted your childwith, and then they leave. So that makes it kind of difficult to regainconfidence in the next teacher because [you] have to go through thatprocess all over again." Teacher turnover, therefore, has a directimpact on some parents, as it affects the relationship between teachersand parents, which may take quite some time to rebuild with a newteacher. Teacher-child relationship: "I'm concerned about mychild's welfare." As a result of this loss of trust, someparents noted concern for their child's well-being. Often, theseconcerns were related to new teachers not being familiar with eachindividual child's needs or any special concerns. One mother wasspecifically concerned about how new teachers would be able to handle anemergency if they did not have background information on the child: Shenoted that "if there was an incident, they are not familiar withthe children in the classroom or what, you know, things they maydo." This parent also noted that because her child was not familiarwith the teacher, he would be less likely to approach the teacher if aproblem arose. Parents also expressed concern that children may beaffected negatively because of changes in teaching style or approach.These changes in teaching were especially salient for parents regardingreactive centers that frequently use substitutes. As one parent noted,"I don't feel that there's any sort of a class program oragenda followed." Thus, parents' concerns for their childrenaddressed not only their physical safety and emotional well-being, butthe quality of the instructional environment as well. It is interesting to note that most of the parents who discussedserious concerns for children's well-being had children who wereenrolled in more reactive centers. Because there was such frequentshifting of teachers in these centers, parents seemed to be less likelyto know who was caring for their children. For example, afterexperiencing two instances of turnover within a few weeks, one parentstated that, "You don't know.., who's taking care of yourchild. You drop him off in the morning and you don't knowwho's going to be there in the afternoon, and that'stough." Another parent noted that the care situation for her childhad changed so frequently that she could not remember the teachers'names. Communication: "Nobody has the history." The majority ofthe parents in both proactive and reactive centers also expressed thesentiment that the loss of relationship with a teacher often resulted inless effective and informative communication. For example, one parentstated, "We feel like there's, um, an interruption.... They[subs or new teachers] can't give you the same kind of informationabout [your child]." Many concerns regarding communication arosefrom new teachers' lack of familiarity with families. One parentsaid, "Nobody has the history ... and the longer term picture ofyour child." Another parent noted that new teachers do not"know me as a parent, they [don't] know what is acceptable tome and what is not acceptable to me in terms of my child'sbehavior." Although parents noted that communication was affectedby turnover, they were less concerned about this, as they understoodthat it takes time to develop new relationships. Specifically, oneparent noted that teachers "are still getting their bearings and sothey don't know what to report at the end of the day." Mostparents mentioned that the communication would return to normal after anadjustment period in which the parents and teachers learned how tocommunicate with a new person. These parents stated, "You have tolearn what that teacher does, how they communicate," and "Ittakes time to develop that kind of relationship again." Trust in relationships: "They probably don't know who totrust." When discussing their children's reactions toturnover, some parents noted that children were more attached tospecific teachers and, thus, were affected negatively by the departureof those teachers. As one parent noted, "When children aresubjected to several different people, it can cause a trust issue."This parent went on to comment that her child had a difficult timebecoming comfortable with the new teacher and expressed concern aboutthe long-term impact that the turnover would have on her child. Anotherparent stated, "I think it has a negative effect on the children,because they don't, urn, again, get to establish a relationshipwith a particular person because there's constant change. So theyprobably don't know who to trust." These sentiments wereechoed by most of the parents, who noticed an effect in their childbecause of the severed relationship with the departing teacher, thuslosing an important person in their child's life. Emotional reactions: "Sad to say goodbye." Some parentsobserved that their children reacted emotionally to teacher turnover.For some children, these reactions conveyed sadness over teachers'departures; as one parent stated, "He came home from school andsaid, 'Mum, we got some bad news today, [the teacher's]leaving.' " Other parents noted similar reactions: "Ithink [she] is, was a little bit sad to say goodbye"; "Ibelieve she was sad." Additionally, many parents noted thatchildren stated occasionally that they missed their previous teacher.However, for other parents, their children showed more intense reactionsof sadness and anxiety about going to school. For instance, a parentstated, "He's been crying like he didn't want togo." Another parent said her child was upset at the transition and"just doesn't want to come [to school]." No matter theintensity of the reaction, there was a consensus among parents thatchildren were sad when they lost their teachers to turnover. Behavioral reactions: "He's been acting out."Parents also observed physical and behavioral reactions from theirchildren. Some parents mentioned that their children expressed theiremotions through acting out, either in the classroom or at home.Specifically, one parent noted that her child was "having liketantrums and stuff.., for the past two weeks." Other parents statedthat their children were less likely to exhibit their normal outgoingbehavior. For instance, one mother said, "I can definitely tellhe's not as comfortable with them, urn, maybe not as willing toparticipate and ... be as outgoing." Similarly, another parentobserved that her child had "become more clingy when being droppedoff at school." Although many of the children's physical orbehavioral reactions were not severe, there were a few parents who notedthat their children had a more difficult time adjusting to the turnover.One mother reported that her child had "wet his bed for like threedays in a row, and he never does ... but he knows what's going onand it's bothering him" Another parent noted that her childhad "a lot of meltdowns.., a lot of crying, not wanting to go tobed, not wanting to come to school" after the turnover started.Although these children expressed their feelings in a variety of ways,the emotional, physical, and behavioral reactions of children indicatethat children and families are impacted negatively by turnover in childcare classrooms. Additionally, although we recognize that some of theseeffects may well be temporary, we must consider the possible long-termimpact of turnover transitions, especially those that are more chaoticor reactive. Quantitative Results Differences in classroom quality from pre-turnover topost-turnover. The quantitative results further capture some of theexperiences discussed by teachers. For example, classroom quality seemsto be compromised during the turnover transition. The overall meanscores on the ECERS-R from the pre-turnover assessment to thepost-turnover assessment decreased (pre M = 4.35; post M = 4.13).Similarly, as reported in Table 4, subscale and factor scorespost-turnover were lower than scores prior to the turnover transition.These data are purely descriptive, as the sample size is small (N = 13)and there is not enough power to test significance. As noted in Table 5,classrooms identified as proactive in their response to turnover were ofhigher quality, based on the ECERS-R scores, than classrooms identifiedas reactive. Additionally, post hoc analyses revealed that reactivecenters experienced a higher turnover rate (29.82%) compared to programsidentified as proactive (20.93%), also reported in Table 5. Teacher-child relationships. Utilizing the STRS, the interactionpatterns and teacher-child relationships for all of the children in eachclassroom with the departing, remaining, and new teachers were examined.This included an examination of the three factors of the STRS scale:Conflict, Closeness, and Dependency. Additionally, the conflict anddependency scores were added to formulate a relational negativity score(Hamre & Pianta, 2001). The findings, also reported in Table 6,indicated that the relationship between teachers and children weredifferentiated by teacher role (departing, remaining, and new).Departing teachers shared a closer relationship with the children in theclassroom, F(2, 302) = 4.17, p = .01, and perceived children to be moredependent on them, F(2, 302) = 14.63, p =.001, as compared to remainingor new teachers. Not surprisingly, this may suggest that the departingteachers know more detailed characteristics of the children inclassrooms that they are leaving, compared to new teachers, andsubsequently indicated a closer and more dependent relationship betweenthe departing teachers and children. DISCUSSION This mixed-methods study on teacher turnover demonstrates thecomplexity of the turnover process and documents its impact on theclassroom environments and people associated with this change. Thequalitative and quantitative findings demonstrate that the turnovertransition affects all individuals involved, as well as the child carecenters as a system. Among these effects are negative changes inclassroom quality, particularly in terms of disruptions to routines andclassroom instruction. Although only descriptive in nature, the changesin the ECERS-R (including overall mean, subscale, and factors scores)indicate that all areas of classroom functioning decrease in qualityduring the turnover transition. Additionally, changes in teacher-parentand teacher-child relationships may compromise trust and rapport, createcommunication difficulties, and induce emotional and behavioraldistress. For example, the STRS data were collected to demonstrate thatthe closest connections between children and teachers were those thatexisted between the departing teachers and the children. Although thismay be inherent in the role of the departing teacher, it is clear that abond has been severed. Furthermore, in this sample, the remainingteacher was often the assistant teacher or a center "floater"who had some level of familiarity with the children but was not asconnected as the departing teacher. Interdependence of Coping According to Bloom (1991), child care systems can be thought of associal systems that include the interdependent components of people,structure, and processes. When change occurs in one component of thesystem, the other components also will experience change. Thus, whenteachers (the people component) change due to turnover, child carecenters and classrooms experience adjustments to the structure andprocesses within the centers. Examples of these changes includeadjustments in communication, teaching practices, classroom environment,and interpersonal relationships. It is evident from the above findingsthat the impact of teacher turnover elicits a range of coping mechanismsfrom directors, teachers, children, and parents in the hopes ofregaining stability in their shared environment. Such coping mechanismsinstigate a dynamic process of teacher turnover that continues toreverberate; how directors, teachers, children, and parents respond toteacher turnover not only changes the nature of their personalexperiences, it also influences their relationships with one another.For example, as a result of having a teacher depart, children may becomesad, resist coming to school, or "test" the new environment.Teachers must then respond not only to an increased workload but also tothe children's varying reactions, while altering the schedule andmodifying the instructional activities in an attempt to manage theclassroom. This subsequently creates more inconsistency for thechildren, resulting in additional stress. As parents respond to thesadness and distress of their child, they also may receive disjointedcommunication from the new teacher, with whom they may be struggling toestablish a new relationship. Prevention From the interviews with directors, teachers, and parents, threemain themes emerged as suggestions to reduce turnover: increasesalaries, offer better benefits, and create supportive workenvironments. Consistently, salary was the number one strategy mentionedby all participants to decrease the rate of turnover. Benefits also werementioned frequently as a way to reduce turnover by professionalizingthe field so that teaching in child care is understood as a long-termcareer rather than as a temporary job. Additionally, various aspects ofthe work environment were discussed, including a supportive and positiveorganizational climate, lower teacher-child ratios, smaller group sizes,and increasing the number of committed teachers with education in earlychildhood. It may be that no single factor is responsible for a teacher'sdecision to leave. Rather, it may be the combination of low salaries,inadequate benefits, and difficult work environments that lead teachersto leave their jobs. This is not news to early childhood researchers andpolicy analysts, but it does provide needed emphasis on the serious,complex, and multilevel nature of the issue. In light of the impact ofturnover on directors, teachers, parents, and children in the child caresystem, it is imperative that we have a goal to drastically reduceturnover by addressing each of these factors. Amelioration of the Impact of Turnover Because prevention will provide only limited assistance with theissue of turnover, programs need to have purposeful strategies in placeto make the transition as smooth as possible. Based on the results ofthe current study, the policies and systems that are implementedproactively by center directors dramatically change the severity of theturnover process. That is, centers that responded with more proactivestrategies that aim to sustain the most consistency for teachers,children, and parents tended to relieve some of the stress associatedwith turnover. Although some reasons for turnover can be positive (e.g.,promotion of a teacher or terminating the employment of a poorlyperforming teacher), establishing policies and procedures to maintainconsistency for teachers, children, and parents is essential to minimizedisruption in the work environment and learning environments of teachersand children. Helping centers create and implement these explicitstrategies to mitigate the negative effects of the turnover transitionmust be a high priority. Taking on Turnover Whitebook and Bellm (1999) provided excellent strategies to assistprograms in coping with turnover. Findings from the current study offerseveral possibilities. For instance, having consistent floaters orsubstitutes who are already familiar with center policies and childrenin the classrooms may prevent teachers or children from being shiftedfrom classroom to classroom when turnover occurs. Other strategies mayinclude having new teachers shadow departing teachers, so that theremaining teachers are not required to spend as much time and energytraining incoming teachers. This would also allow children to form arelationship with the new teacher prior to the turnover of the departingteacher. Additionally, maintaining consistent mechanisms ofcommunication with parents seems vital as parents form new relationshipswith new teachers. Having proactive strategies allowed centers to handlethe turnover transition in a way that supported children, families, andteachers by minimizing the disruption associated with the turnover.Thus, these centers were able to maintain high-level interactions andprevent drastic decreases in classroom quality. Nonetheless, proactiveprograms were not immune to turnover. Proactive and reactive programsexperienced turnover, and the causes of turnover including low salaries.Although the proactive centers reported a lower turnover rate, it isimportant to note that the differences between these two types ofprograms were how they coped with turnover, not that one typeexperienced turnover and the other did not. Limitations The results of the current study highlight the complexity of theteacher turnover process. Nevertheless, there were a few limitations inthe current study. Unfortunately, the major limitation of the currentstudy was its small sample size. It was extremely difficult to solicitcooperation from programs when they were experiencing turnover. Turnoveris a time when most programs do not want observers in the classrooms andstress in the facility is high. We are deeply grateful to the programsthat shared their turnover experiences with us. However, the smallsample lowered the power necessary for testing statistical significanceon changes in classroom quality and restricted which analyses could beconducted. Additionally, as the sample was not randomly selected, it isnot representative of all child care and does not allow forgeneralizations to be made. The findings are also limited by thesemistructured nature of the interviews. Such structure may havehindered participants from voicing concerns about, or experiencesrelated to, the turnover transition that were not addressed with thepreset questions. CONCLUSION The study allowed for measurement of the immediate impact ofturnover rather than the retrospective findings that have been reportedin past research. Teachers were interviewed as they were leavingcenters, arriving for employment, or experiencing the loss of acoteacher. Directors and parents were caught in the turmoil of theturnover crisis. We believe that because of the methodology used, wecaptured the day-to-day reality of turnover, including the stress andconcern created. Furthermore, parents, teachers, and directors were ableto provide candid and accurate responses, because the impact of teacherturnover was very fresh. Future research certainly should include alarger sample size with a similarly immediate impact format. Theturnover variable also should be included in large-scale, longitudinalstudies that include child outcome data. One important lesson learned is that good things do seem to gotogether. Proactive centers may be able to better minimize the impact ofturnover on classrooms, teachers, and families, because they also havesuccessful management strategies and demonstrate an ability to maintainhigher quality, even through the disruption of turnover. Furtherexamination of the strategies used by proactive centers will enable usto assist in better preparing programs for turnover and perhaps todecrease the level of turnover in our child care facilities. Preparingprograms to reduce and minimize the effects of teacher turnoveralleviates job stress among teachers and directors, while providing amore positive experience of child care for children and parents. DOI: 10.1080/02568543.2011.533118 APPENDIX Teacher Interview Questions (adapted from Center for the Child CareWorkforce, 2000) 1. In the last year, or since you have been working at this center(if less than one year), how many changes of either type have youexperienced? a. Number of changes in groups of children b. Number of changes in rooms 2. What accounts for these changes? 3. What is the total number of teachers assigned to work in yourclassroom each day? 4. During the last year, how many different teachers have youworked with in your classroom? 5. Do the teachers you work with consistently share the workloadwith you on a regular basis? 6. Have you received as much training as you would like (in thelast year)? a. If no, please indicate why not. 7. If you leave this job, what would you be most likely to do? 8. When staff leave your center or when there are staff vacancies,does it affect your ability to do your job? a. On a scale of 1 to 10, how difficult is it for you? b. What areas of your teaching are most affected? 9. When staff in your center are out of the classroom for a fewhours, how does it affect your ability to do your job? A day? A fewdays? a. Please explain how. b. On a scale of 1 to 10, how difficult is it for you? c. What areas of your teaching are most affected? 10. Does turnover among coworkers affect your own career goals? 11. How long do you plan to remain in child care? 12. Would you recommend teaching in child care as a career choice? a. Please explain. 13. What are the most rewarding aspects of your job? 14. What are the most stressful aspects of your job? 15. (If a departing teacher) Why are you leaving this job? 16. (If a new teacher) Why did you accept this position? 17. How do you think children are affected when a teacher is out ofthe classroom for a short time? a. Describe their behavior. 18. How do you think children are affected when a teacher resignsor is fired? a. Describe their behavior. 19. How do you think parents feel when teachers leave? 20. How do you think parents feel about substitute teachers in theclassroom? 21. How do you think things will change next week when (departingteacher) leaves? REFERENCES Bloom, P. J. (1991 ). Child care centers as organizations: A socialsystems perspective. Child and Youth Care Forum, 20, 313-333. Bloom, P. J. (1997). Navigating the rapids: Directors reflect ontheir career and professional development. Young Children, 52(7), 32-38. Bureau of Labor Statistics. (2007). May 2007 national occupationalemployment and wage estimates. Retrieved fromwww.bls.gov/oes/current/oes_nat.htm. Cassidy, D. J., Hestenes, L. L., Hegde, A., Hestenes, S., &Mims, S. (2005). Measurement of quality in preschool child careclassrooms: An exploratory and confirmatory factor analysis of the earlychildhood environment rating scale-revised. Early Childhood ResearchQuarterly, 20, 345-360. Center for the Child Care Workforce. (2000). Child care staffsalary and working conditions survey. Washington, DC: Author. Center for the Child Care Workforce. (2004). Current data on thesalaries and benefits of the U.S. early childhood education workforce.Washington, DC: Author. Creswell, J. W. (2005). Educational research: Planning, conducting,and evaluating quantitative and qualitative research. Upper SaddleRiver, NJ: Pearson. Curbow, B., Spratt, K., Unagretti, A., McDonnell, K., &Breckler, S. (2001). Development of the child care worker job stressinventory. Early Childhood Research Quarterly, 15, 515-535. De Kruif, R. E. L., McWilliam, R. A., Ridely, S. M., & Wakely,M. B. (2000). Classification of teachers' interaction behaviors inearly childhood classrooms. Early Childhood Research Quarterly, 15,247-268. DeVita, C., Twombly, E., & Montilla, M. (2002). Toward betterchild care worker compensation: Advocacy in three states. (ERIC DocumentReproduction Service No. ED464715). Glaser, B. G., & Strauss, A. L. (1967). The discovery ofgrounded theory: Strategies for qualitative research. New York: Aldinede Gruyter. Goodman, I., Brady, J., & Desch, B. (1987). A commitment toquality: The impact of state supplemental funds on Massachusetts HeadStart. Newton, MA: Education Development Center. Hale-Jinks, C., Knopf, H., & Kemple, K. (2006). Tacklingteacher turnover in child care: Understanding causes and consequences,identifying solutions. Childhood Education, 82, 219-226. Hamre, B. K., & Pianta, R. C. (2001). Early teacher-childrelationships and the trajectory of children's school outcomesthrough eighth grade. Child Development, 72(2), 625-638. Harms, T., Clifford, R. M., & Cryer, D. (1998). Early childhoodenvironment rating scale-revised. New York: Teachers College Press. Harms, T., Cryer, D., & Clifford, R. M. (2006). Infant/toddlerenvironment rating scale-revised. New York: Teachers College Press. Howes, C., & Hamilton, C. E. (1993). The changing experience ofchild care: Changes in teachers and in teacher-child relationships andchildren's social competence with peers. Early Childhood ResearchQuarterly, 8, 15-32. Howes, C., Hamilton, C. E., & Philipsen, L. C. (1998).Stability and continuity of child-caregiver and child-peerrelationships. Child Development, 69, 418-426. Howes, C., Phillips, D. A., & Whitebook, M. (1992). Thresholdsof quality: Implications for the social development of children incenter-based child care. Child Development, 63, 449-460. Howes, C., & Smith, E. W. (1995). Relations among child carequality, teacher behavior, children's play activities, emotionalsecurity, and cognitive activity in child care. Early Childhood ResearchQuarterly, 10, 381-404. Manlove, E. E., & Guzell, J. R. (1997). Intention to leave,anticipated reasons for leaving, and 12-month turnover of child carecenter staff. Early Childhood Research Quarterly, 12, 145-167. Mertens, D. M. (1998). Research methods in education andpsychology: Integrating diversity with quantitative and qualitativeapproaches. Thousand Oaks, CA: Sage. North Carolina Department of Health and Human Services, Division ofChild Development. (2010). The NC star rated license: Overview.Retrieved from http://ncchildcare.dhhs.state.nc.us/parents/pr_sn2_ov_sr.asp Olenick, M. (1986). The relationship between quality and cost inchild care programs. Unpublished doctoral dissertation. UCLA Center forChild and Family Policy Studies, School of Social Welfare, University ofCalifornia at Los Angeles. Phillips, D., Mekos, D., Scarr, S., McCartney, K., &Abott-Shim, M. (2000). Within and beyond the classroom door: Assessingquality in child care centers. Early Childhood Research Quarterly, 15,475-496. Pianta, R. C. (2001). Student-Teacher Relationship Scale (STRS)professional manual. Odessa, FL: Psychological Assessment Resources. Rennie, D. L. (2006). The grounded theory method: Application of avariant of its procedure of constant comparative analysis topsychotherapy research. In C. T. Fischer (Ed.), Qualitative research:Instructional empirical studies (pp. 59-78). New York: Elsevier. Stremmel, A. J. (1991 ). Predictors of intention to leave childcare work. Early Childhood Research Quarterly, 6, 285- 298. Whitebook, M., & Bellm, D. (1999). Taking on turnover: Anaction guide for child care center teachers and directors. Berkeley, CA:Center for the Child Workforce. Whitebook, M., & Sakai, L. (2003). Turnover begets turnover: Anexamination of job and occupational instability among child care centerstaff. Early Childhood Research Quarterly, 18, 273-293. Whitebook, M., Sakai, L., Gerber, E., & Howes, C. (2001). Then& now: Changes in child care staffing, 1994-2000. Washington, DC:Center for the Child Care Workforce. (ERIC No. ED452984) Deborah J. Cassidy, Joanna K. Lower, and Victoria L. Kintner-Duffy The University of North Carolina at Greensboro, Greensboro, NorthCarolina Archana V. Hegde East Carolina University, Greenville, North Carolina Jonghee Shim Auburn University, Auburn, Alabama Submitted May 18, 2009; accepted July 15, 2010. Address correspondence to Deborah J. Cassidy, Department of HumanDevelopment and Family Studies, The University of North Carolina atGreensboro, 248 Stone Building, P.O. Box 26170, Greensboro, NC 27402.E-mail: djcassid@uncg.eduTABLE 1Frequencies for Teacher Demographics Teacher Status All Departing Remaining New (N = 34) (n = 11) (n = 10) (n =13)Highest level of education High school/General 3 1 0 2 Equivalency Diploma Some college 15 4 5 6 2-year degree 3 1 1 1 4-year degree 2 2 0 0 Some graduate school 3 1 1 1 Graduate degree 1 0 1 0 Missing 7 2 2 3Ethnicity African American 9 3 3 3 European American 17 6 4 7 Other 1 0 l 0 Missing 7 2 2 3TABLE 2Employment Information by Teacher Status (Means and SDs) Teacher Status All DepartingTime employed in child care (in 78.39 (70.01) 80.75 (51.71) months)Time employed at current center 25.05 (27.52) 32.38 (30.44)(in months)Hourly wage $8.82 ($1.94) $8.73 ($0.98)Benefits received from center Health coverage None 8 2 Partial 6 3 Full 11 3 Retirement package No 23 6 Yes 3 2 Teacher Status Remaining NewTime employed in child care (in 108.71 (101.7) 59.53 (57.66) months)Time employed at current center 38.71 (26.29) 11.43 (22.08)(in months)Hourly wage 9.71 ($2.29 $8.50 ($2.28)Benefits received from center Health coverage None 2 3 Partial 0 3 Full 5 3 Retirement package No 7 9 Yes 0 1TABLE 3Frequencies for Director DemographicsHighest level of education 2-year degree 2 4-year degree 3 Some graduate school 3 Graduate degree 1Ethnicity African American 2 European American 6 American Indian 1TABLE 4Mean Differences in Classroom Quality Pre- and Post-Turnover Pre-Turnover Mean Post-Turnover Mean (SD) (SD)ECERS-R Average Score 4.35 (l.02) 4.13 (l.29)ECERS-R SubscalesSpace and Furnishings 4.57 (.86) 4.40 (l.18)Personal Care Routines 3.49 (l.26) 3.39 (l.46)Language and Reasoning 4.66 (.95) 4,18 (l.31)Activities 4.25 (l.06) 4.16 (l.31)Interactions 4.88 (l.76) 4.43 (l.88)Program Structure 4.77 (l.63) 4.14 (l.50)ECERS-R FactorsActivities/Materials 3.83 (l.08) 3.74 (l.22)Language/Interactions 4.82 (l.55) 4.29 (l.64)ECERS-R = Early Childhood Environment Rating Scale-Revised.TABLE 5Mean Quality Ratings Based on Reactivity to Turnover Proactive Mean (SD) Reactive Mean (SD)ECERS-R Average Score 5.11 (1.07) 3.61 (.03)ECERS-R Subscales Space and Furnishings 5.23 (.95) 3.96 (.68) Personal Care Routines 4.21 (1.70) 2.89 (.61) Language and Reasoning 5.32 (.93) 3.78 (.81) Activities 5.00 (1.27) 3.63 (.67) Interactions 6.08 (.85) 3.64 (1.59) Program Structure 5.08 (l.78) 4.01 (1.28)ECERS-R FactorsActivities/Materials 4.44 (l.15) 3.32 (.87)Language/ Interactions 5.82 (.99) 3.65 (l.28)ECERS-R = Early Childhood Environment Rating Scale-Revised.TABLE 6STRS Factors by Teacher Classification Departing Remaining New (N = 128) (N = 85) (N = 92)Closeness 4.08 * (-0.62) 3.87 (-0.65) 3.86 (-0.67)Conflict 2.03 (-0.89) 2.13 (-0.61) 2.01 (-0.86)Dependency 2.45 ** (-0.87) 2.32 (-0.74) 1.83 (-0.93)Relational negativity 2.13 (-0.76) 2.17 (-0.57) 1.97 (-0.77)* p < .01, ** p = .001.