Arizona State University Social Networking Postings: Views from School Principals

Numerous recent media accounts indicate that teachers are being fired, put on probation, or otherwise censured because of information found on their social networking sites (SNS). While the literature in business, psychology, and pharmacy shows initial investigations of the impact of SNS information on hiring decisions, this area has not been investigated in the field of education. Data from pre-service teachers' SNS were compiled into a 51-item questionnaire and K-12 school principals rated each statement on its likelihood to influence a hiring decision if it were found on the SNS of a teacher applying for a position at his or her school. The statements included on the questionnaire included sexual talk, swearing, violent language, references to drugs and/or alcohol use, and language that could be considered pejorative to individuals with disabilities, persons of color and/or homosexual individuals. The findings of this preliminary study indicate that of the 51 statements, 28 had average ratings in the moderate-to-significant impact on hiring decision range and 22 had average ratings in the minor-to-moderate impact on hiring decision range. Only one item of the 51 fell in the no impact-to-minor impact on hiring decision range. Findings are discussed in relation to professional dispositions, categories of information that may influence hiring decisions, and internet use policies. Implications based on the data from this study include the need to develop state and university policies for SNS use. Finally, the results of this study indicate that further research into hiring administrators' use of data from SNS in hiring decisions is warranted.

Anúncios nas redes sociais: Opiniões dos/as Diretores/as de Escolas Resumo: Numerosos relatos recentes da mídia indicam que os professores estão sendo demitidos, postos em liberdade condicional, ou censurados por causa de informações encontradas em seus sites de redes sociais (SNS).Embora a literatura em negócios, psicologia, farmácia e mostra as investigações iniciais sobre o impacto das informações SNS em decisões de contratação, esta área não tem sido investigada no campo da educação.Os dados de pré-serviço de professores SNS foram compilados em um questionário de 51 itens e K-12 diretores de escolas avaliado cada declaração sobre sua probabilidade de influenciar uma decisão de contratação, se foram encontradas no SNS de um professor aplicando para uma posição ou em sua sua escola.As declarações incluídas no questionário incluiu conversa sexual, palavrões, linguagem violenta, referências a drogas e / ou uso de álcool, e uma linguagem que poderia ser considerado pejorativo para pessoas com deficiência, pessoas de cor e / ou indivíduos homossexuais.Os resultados deste estudo preliminar indicam que dos 51 depoimentos, 28 tinham avaliações médias no impacto moderado a importante decisão sobre a contratação de gama e 22tiveram classificação média no impacto menor a moderada na contratação de gama decisão.Apenas um item dos 51 caiu na nenhum impacto impacto-a-menor sobre a contratação de gama decisão.Os resultados são discutidos em relação às disposições profissionais, categorias de informações que podem influenciar as decisões de contratação, e políticas de uso da internet.Implicações com base nos dados deste estudo incluem necessidade de desenvolver políticas estaduais e universidades para uso SNS.Finalmente, os resultados deste estudo indicam que Introduction Some commentators suggest that the private lives of teachers may be under more scrutiny than those in other professions (Bock, 2008;Vacca, 2005).Some states, including Florida, Massachusetts, Colorado, Tennessee, Georgia, Iowa, Michigan, and North Carolina have removed, reprimanded, or suspended teachers for their social networking site (SNS) postings (e.g., Associated Press, 2010;cbsatlanta.com, 2009;Geha & Sterling, 2009;Hui, 2010;Shapira, 2008).Brief perusals of SNS and recent publications (e.g., Bock, 2008;Peluchette & Karl, 2010;Read, 2007) indicate that college students post information that may not be considered consistent with a professional persona and may be objectionable to future employers, state boards of education, and accrediting bodies.Such information may include sexual references, vulgar language, violent talk, or discriminatory statements.The purpose of this study is to examine the possible influence that such information posted on SNS by teaching position applicants might have on the hiring decisions of principals.We begin by situating this problem in discussions of teacher professional behavior and dispositions, SNS and hiring practices, teacher use of SNS, and legal issues such as privacy and conduct unbecoming a teacher.

Teacher Professional Behavior and Dispositions
Numerous professional bodies propose standards or codes of behavior for teachers.The National Council for Accreditation of Teacher Education (NCATE) is a primary accreditation organization for colleges, schools, and departments of education, currently accrediting 657 colleges of education, with 100 more seeking accreditation (about NCATE, p.1, 2010).NCATE proposes to ensure higher quality teacher preparation and to maintain higher quality educational practices through its performance-based professional accreditation standards.Standard 1g indicates: Candidates for all professional education roles develop and model professional dispositions that are expected of educators… Professional dispositions are not assessed directly; instead the unit assesses dispositions based on observable behavior in educational settings {emphasis added} (paragraph 9, Supporting Explanation, Standard 1, NCATE, 2007).Similar statements related to professional teacher dispositions and behavior have been issued by the Interstate Teachers Assessment and Support Consortium (InTASC); (CCSSO, 2010, p.19) and the Georgia Professional Standards Commission (GaPSC) in the Code of Ethics for Educators (GaPSC, 2009, p. 4), to name a few.There is disagreement over a precise definition of the term "dispositions" (e.g., Singh & Stoloff, 2007;Stooksberry, Schussler, & Bercaw, 2009;Thornton, 2006;Welch, Pitts, Tenini, Keunlen & Wood, 2010) and the meaning of morality as applied to teaching (Burant, Chubbuck, & Whipp, 2007;Campbell, 2008;Lumpkin, 2008;Osguthorpe, 2008).However, many colleges of education and state codes of ethics still include an expectation that teachers will exhibit professional dispositions and behavior and will avoid moral turpitude or behavior unbecoming a teacher (e.g., ETSU COE, 2010;Fischetti, J., Imig, S., Ndoye, A., & Smith, R., 2010;GaPSC, 2009;Pennsylvania Department of Education, 2011;Professional Teaching Practices Commission [AK], 2000; State Board for Educator Certification [TX], n.d.; Union County [NC] Board of Education, 2011).

SNS and Hiring
Corporate employers are increasingly using SNS to search for information on prospective employees (Alfonsi, 2006;Amburgey, 2006;Bock, 2008;Fuller, 2006).An officer of a major outplacement company indicates that virtually all recruiting firms use SNS as recruiting grounds (Jannsen, 2009), and other sources confirm that employers do seek information on potential employees by looking at their SNS (Bergstrom, 2008;CCH HR Management, 2009;Telegraph Media Group, 2010).Simply having a SNS profile is not likely to cost one a job; it is what is in the profile that counts.Sexually explicit information, information on drug use, illegal behavior, drunkenness, a poor work ethic, racial or sexist comments, comments about past employers or coworkers, photos showing partying or scant clothing -can all cause employers to reconsider an intent to hire (Alfonsi, 2006;Bock, 2008;Bohnert & Ross, 2010;Fuller, 2006;Telegraph Media Group, 2010).In a 2009 CareerBuilder survey of 2,600 hiring managers, 45% of respondents (an increase from 22% in 2008) indicated that they use social networking sites to screen job applicants.Thirtyfive percent of respondents found information on these sites that caused them not to hire an applicant (Grasz, 2009).
While the CareerBuilder survey found a 104% increase in hiring managers' use of SNS to screen job applicants from 2008 to 2009 (Grasz, 2009), a number of other non-scholarly sources describe employment practices that include looking at SNS, such as Reader's Digest (Simmons, 2008), The New York Times (Finder, 2006, June 11), Nursing Standard (Royal College of Nursing, 2009), and Computerworld (Havenstein, 2008, September 12).Scholarly sources in business have also published studies examining the use of SNS in hiring decisions (Elzweig & Peeples, 2009;Oleniczak, Pike, Mishra, & Mishra, 2010), cautioning job-seekers against believing that SNS are private and offering recommendations for professional self-presentation.A review of the issues surrounding online SNS and the implications for pharmacy education suggests that these SNS pose a danger to student privacy and professional reputation, and the author recommends that schools take steps to educate students about these dangers (Cain, 2008).
The few empirical studies on how employers respond to content posted on SNS have been published in business and psychology journals and have examined topics such as the purposes for which individuals use SNS (Agarwal & Mital, 2009), and students' intended images on Facebook (Peluchette & Karl, 2010).More directly related to the purpose of the current study, Kluemper and Rosen (2009) examined whether applicant profile information found only on SNS could be used to distinguish high and low scorers on personality traits that are used to improve employment selection decisions.They found that SNS information could be used by employers to accurately make employment decisions.Similarly, Bohnert and Ross (2010) examined the influence of information found on SNS on others' evaluations of prospective employees and found that applicants with a professional-oriented or family-oriented SNS presence were considered more desirable for hire than those with an alcohol-oriented SNS.In a study of SNS use and appropriateness of content, respondents generally indicated they were neutral about employers or strangers viewing their SNS profile, but about 20% indicated they had information in their profile they would not want current or prospective employers to see (Peluchette & Karl, 2008).While research on SNS and hiring practices is beginning to appear in some fields, the educational literature is noticeably void of research on this topic.

SNS, Teacher Education and Law
Media accounts of teacher censure by educational institutions abound, indicating that persons other than SNS "friends" are looking at SNS postings (e.g., Associated Press, 2008b;Associated Press, 2010;Geha & Sterling, 2009;Hui, 2010;Read, 2007;WKOWTV.com, 2009).One of the first accounts describes Stacy Snyder, who was planning to graduate in 2006 from Millersville University of Pennsylvania with a degree in education -until a picture of her wearing a pirate hat and drinking from a plastic cup with the caption "Drunken Pirate" appeared on MySpace.Millersville University administrators found the picture unprofessional and Ms. Snyder was denied her degree in education (and, consequently, her teaching certificate) and awarded a degree in English instead (Read, 2007).
Ms. Snyder (Read, 2007) is not the only teacher to have been censured for inappropriate SNS content.In the Charlotte (NC)-Mecklenburg school district, several teachers faced disciplinary action for posting a variety of material that administrators found objectionable.One teacher wrote "I hate my students!"and another was fired for referring to her students as "chitlins" and her school as "the most ghetto school in Charlotte" (Associated Press, 2008b).During a trip to Europe, a high school teacher from Barrow County, GA was photographed several times with a glass of wine or beer in front of her.These photos, along with an expletive attributed to the teacher, were posted on her private Facebook page and she was asked to resign from her teaching position (WSBTV.com, 2009;cbsatlanta.com, 2009).In Austin, TX, art teacher Tamara Hoover was escorted from her high school classroom and termination procedures were initiated after nude pictures of her were found on Flickr, a photo-sharing site (May, 2006).These are only a few of the cases in which teachers have been reprimanded for postings on their SNS.
At Georgia Southern University (GSU), students are introduced to the notion that SNS postings, among other behaviors, may not be consistent with professional teacher behaviors.This introduction to professional teacher behavior and dispositions occurs in the Pre-Professional Block (PPB).All pre-education majors seeking a Bachelor of Science in Education degree and initial teacher certification at GSU must complete the PPB, a series of three courses and a 51-hour observation-based field experience.The purpose of the PPB is to: (1) provide university students with a realistic picture of the workings of a school and the day-to-day responsibilities of a teacher; (2) encourage students to develop observation, reflection, and critical thinking skills; and (3) serve a "gatekeeping" function for the Teacher Education Program (TEP).Students must receive a satisfactory evaluation in the Professionalism components of the practicum prior to admittance to TEP (Griffin, 2010;Griffin, deMaille, Lake, & Hotchkiss, 2008).Professionalism focuses on the entry-level dispositions and behaviors of potential teacher candidates, including attendance, appropriate attire, punctuality, non-disruptive behavior, serving as a respectful role-model, demonstrating professional interactions in school settings, and overall impressions of the teacher candidate.
Recently, some of the teacher candidates at GSU have come under scrutiny for nonprofessional behavior related to their online information.For example, a student teacher spent one morning of her student teaching placement removing pictures of herself in a bikini from secondary school corridors -the photo was obtained from a SNS by one of the high school students.A second student was removed from his field placement site for leaving writings of a sexually explicit nature at the school.In his second attempt at completing PPB practicum, he invited students to "friend" him on Facebook.Examination of his Facebook page revealed postings containing vulgar language, discriminatory language, and sexually explicit language.The student was subsequently removed from his field placement for a second time.A third student was repeatedly late and was admonished to be on time to his field placement by both the Clinical Supervisor (CS) and the PPB Coordinator (2005).The CS found the student's profile on Facebook and discovered a "joke" about child molestation on the site; the student was immediately removed from the field at the request of the host school.Researchers at Arizona State University also report that at least one student teacher has been dismissed from an internship because of information posted on his MySpace site (Foulger, Ewbank, Kay, Popp & Carter, 2009).
As a result of the incidents described above and others that have occurred at GSU, a Statement on Internet Social Network and BLOG Sites (SNS Statement) has been adopted for all teacher education program practicum experiences.At PPB orientation sessions, this policy is introduced and a faculty member provides a lengthy discussion of materials appropriate for posting on SNS.Students are strongly advised to remove any photographs or descriptions of drug or alcohol use, sexually explicit material, and anything else inconsistent with GSU and GaPSC, InTASC and NCATE expectations of professionalism.While GSU does not routinely monitor SNS, the SNS Statement provides for immediate removal from a field placement at the request of the host school if objectionable material is found on the student's SNS by faculty at the host school or GSU.
When SNS, such as Facebook (the SNS commonly discussed at GSU), are discussed in PPB orientation sessions or classes, students have claimed that their Facebook is set to "Private" and therefore not viewable by others unless they are invited to view it.Students at GSU have also claimed that, in general, their Facebook is private and it is a violation of their privacy to look at it if not invited -despite its very public appearance on the Internet.In fact, it is easy to access information on Facebook, either by issuing a friend request to another Facebook user, or by viewing information through another user's site.For example, Person A could see Person B's information (even though Person B rejected A's Facebook friend request) by physically looking at Person C's account, who is a mutual friend of A and B. It is virtually impossible to know who is viewing a Facebook or MySpace site.
While the right to privacy is not a right explicitly guaranteed by the Constitution of the United States (Drake v. Covington County Board of Education, 1974) there exist areas or zones of privacy under the Constitution in the First and Fourth Amendments.In their pioneering paper, Warren and Brandeis (1890) recognized a need for the right to privacy as a remedy to the circulation of photographs and invasion of privacy by newspapers.Years later in a landmark legal case (Katz v. United States, 1967), the Supreme Court held that people have a reasonable expectation for privacy in certain situations.By 1999 Sun Microsystems' CEO Scott McNealy, commenting on the profusion of information available on computerized databases, made the observation, "You have zero privacy anyway.Get over it" (Sprenger, 1999).Consistent with McNealy's somewhat cynical view, Williamson (2009) noted that conventional wisdom acknowledges that information posted on the Internet becomes public information.The California appellate court confirmed this view in a recent ruling.Cynthia Moreno published a negative essay about her hometown on her MySpace blog and after a newspaper published it and her parents were harassed, she filed an invasion of privacy suit.The appellate court ruled that once she placed her thoughts online, she relinquished her right to privacy.Similarly, "[employers] can feel comfortable that if an employee posts something to Facebook or MySpace, that's considered a public disclosure" (Neuberger, in Williamson, 2009, p. 67).Bick (2010) suggests the Internet has had the effect of changing a person's expectation of the right to privacy and has consequently reduced privacy rights.In Romano v. Steelcase, Inc. (2010), the New York Supreme court lent credence to this claim by ruling that it was unreasonable to expect privacy on a SNS since the purpose of these sites is to share information.
A challenge in legal decisions related to teachers and other public employees has been assessing the rights of public employees to free speech and privacy.Marvin Pickering (Pickering v. Board of Education, 1968) was fired after a letter he wrote criticizing the School Board was published in a local newspaper.Pickering was reinstated, and this case led to the Pickering balancing test for evaluating the speech of public employees.In essence, this balancing test comes into play if an employer takes action against a public employee for something stated publicly.The question is whether or not the employer's action violates the employee's First Amendment right to freedom of speech and the balancing test requires an examination of whether the speech under question is clearly tied to a matter of public concern and whether it undermines the functioning of the public employer.The Pickering balancing test was applied in Melzer v. Board of Education of the City School District of the City of New York (2003).Melzer was terminated after a television station aired footage of him counseling a new member at a North American Man Boy Love Association (NAMBLA) meeting.A stated goal of NAMBLA is to promote legalization of sexual relationships between men and boys, and Melzer's membership and speech in support of NAMBLA was considered to be a matter of public concern as related to his role as a high school teacher.
Because of expectations from the public, the employer, and state certification commissions, the off-campus behavior of teachers is important (Weldon, 2003) and there are expectations for what teachers ought to do and ought not to do (DiCenso, 2005).Consequently, a number of legal cases have centered on examining conduct unbecoming a teacher, for behavior both in school and out of school.For example, in New York a teacher was dismissed for ten traffic violations, including one driving under the influence of alcohol, and for calling in sick when he was not ill.He was found guilty of conduct unbecoming a teacher, a decision overturned for the traffic violations and upheld for calling in sick, by an appellate court (Board of Education of the Warsaw Central School District, 1994).In Dixon v. Clem (2007), David Dixon was dismissed from his teaching position for conduct unbecoming a teacher after it was made known that he took photos of a female student who was nude from the waist up.Similarly, after three misdemeanor charges and one felony burglary charge, all following the end of a sixteen year romance, Carolyn Hutchison was dismissed "because she had engaged in conduct which rendered her unable to be a role model for her students" (Hutchison v. Kentucky Unemployment Insurance Commission, 2010, para.2).Finally, Pamela Mann was dismissed for falsifying records after repeatedly recording incorrectly her arrival time at work and was charged with conduct unbecoming a teacher (The Gallatin County Board of Education v. Mann, 1998).
The discussion above highlights how teachers can lose their jobs for various types of conduct unbecoming a teacher.Likewise, teachers have been censured for postings on their personal SNS (e.g., Associated Press, 2008b;May, 2006;Read, 2007).It is clear that employers in the corporate sector view SNS when making hiring decisions (e.g., Alfonsi, 2006;Amburgey, 2006;Bock, 2008;Fuller, 2006;Grasz, 2009).However, what is not clear is school principals' views about information posted on SNS by prospective hires.The purpose of this descriptive study is to explore the influence that postings on SNS might have on the hiring decisions of school principals.

Method Participants
School principals were targeted as participants for this study because they are responsible for teacher hiring recommendations provided to district Boards of Education in Georgia.Georgia is divided into 17 Regional Educational Service Area (RESA) districts which are geographically bounded areas within the RESA organization, an organization developed to improve educational services throughout the state of Georgia (Georgia Department of Education, 2005).To select school principal participants, the school systems in the 17 Georgia RESA districts were alphabetized within each district.RESA districts were used as school system clustering units in an attempt to assure better representation of all geographic areas in the state.A random number generator (2009, www.random.org) was used to identify three school systems within each RESA district (except Metro Atlanta) from which participation would be sought using a stratified random sampling method.Five districts in the Metro Atlanta area were identified in an attempt to obtain permission from a district in that geographic region.In all, permission to collect data was requested of 53 school districts (3 each from 16 RESA districts and 5 from the Metro [Atlanta] district) with 16 school districts granting permission to conduct research.Of the remaining 34 school districts contacted, 22 did not respond to repeated requests (i.e., initial phone calls to determine who was the correct contact person, a call to that person, and two emails to document/confirm permission to collect data); six denied permission; the application deadline for requesting permission was passed in two systems; and in four systems, permission was granted but repeated attempts to secure current principal email addresses did not yield the addresses promised.In the 16 school systems from which data were collected, a total of 94 principals were asked to respond to the instrument.Forty-three (45.7%) responded, with 39 (41.5%) providing data other than demographic information only.
The participating school systems represent a wide geographic range within the state of Georgia.While north Georgia was not well-represented (3 of the non-participating districts were in the northernmost part of the state), the Griffin district in the northeast corner did have one participating school system.None of the school systems in the metropolitan areas (e.g., Atlanta, Augusta, Savannah, Macon) granted permission to the research team to conduct the study.
Demographic data for 39 principals ranging from preschool to high school are presented in Table 1.Respondents included 21 males, 18 females, 30 Caucasians and eight African-Americans, as well as one participant who did not provide race or sex data.Twenty-nine (74.4%) respondents were between 36 and 55 years old, 28 (71.2%)respondents had ten years or less experience as principal.The most recent Georgia principal demographic data (Afolabi, Nweke, Stephens, & Toth, 2003) suggest that the sample respondents were fairly representative of the Georgia principal population.The biggest discrepancy between this sample and the Georgia population was in the sex of the respondents.Afolabi, Nweke, Stephens, and Toth (2003) found that 55.1% of principals were female compared to 46.2% in this study; 71.2% were White, compared with 76.9% in this study; and finally, 64.1% of respondents in this study hold the Ed.S. degree as compared with 66.3% statewide in 2003.However, since there were no respondents from metropolitan areas of the state, the results may not generalize to principals in metropolitan areas of Georgia or other states.

Instrument and Procedures
An effort was made to establish content validity of the questionnaire used in this study by examining the SNS pages of students enrolled in an initial education practicum course.These pages provided real content that was produced by potential teacher applicants, and this content was utilized as the source of the questionnaire content.Students' MySpace and Facebook sites with minimal privacy settings were examined by a team of graduate students.All pages examined were accessible by a simple search of the respective social network site; no students were "friended" for the purpose of examining sites, nor was any information obtained through any other form of deception.Comments drawn from the sites were quoted directly as presented by the teacher applicants on the SNS.To protect the identities of the students whose SNS were sampled, descriptions of photographs rather than actual images were used at all stages of the project.Information related to alcohol or drug use, discriminatory or sexually explicit language, cursing, or photos depicting partying behavior or scant clothing were copied into a database.Example database entries include: a) Im gon punch u in ur throat; b) B----!I might be.; c) Turning 21 comes with the responsibility of legal binge drinking; d) photo description: Two females in front of GA state flag with 10 solo cups on table in front of them and caption: BEER PONG CHAMPS!!!This information was used as the basis for the questionnaire because it is the type of information described in media accounts of teachers censured for SNS postings and it is inconsistent with the ethical standards held for teachers by organizations such as the Georgia Professional Standards Commission (2009).All verbatim statements and photograph descriptions in the database were compiled into an instrument for identifying school principals' evaluations of material on SNS.
The instrument was field tested with college faculty for clarity of wording, layout and ease of responding.Fourteen faculty members were invited to provide feedback, and nine responded.On the basis of faculty feedback, minor changes were made to the questionnaire instructions and the layout of the instrument.The next step of field testing involved a random selection of principals throughout the state of Georgia.A list of principal email addresses (2,497 usable addresses) was obtained from the Department of Education, and every 25 th principal on the list was sent an e-mail request to participate in the pilot study of the instrument.In all, 102 principals were asked to respond and 18 completed the instrument, for a response rate of 17.64%.All respondents indicated the instrument was easy to read and complete and no changes were made to the questionnaire.*Respondents in the Other category included one respondent (2.3%) each in the following settings: middle/high school (6-12); alternative school (6-12); higher education; central office The instrument was administered electronically online to study participants. 1Administrators were asked to identify for each quotation or description of a photo from a SNS to what degree, if 1 Initially, the researchers hoped to randomly select principals to participate in the study, as was done for the field test of the instrument.However, during field testing, it was brought to the researchers' attention that many districts in the state would not allow their employees to participate in any research endeavors without permission from the district office.Thus, stratified random sampling by RESA district was used to make obtaining district permission more manageable.Institutional Review Board guidelines for informed consent and voluntary participation were followed.In those systems for which permission to administer the questionnaire was obtained, administrators were sent three e-mails over a six-week period asking them to respond to the questionnaire.No emails were returned due to bad addresses.
any, the information provided would influence a hiring decision if the principal found the information on the SNS of a potential teacher.Principal evaluations were based on a four-point Likert scale, with a rating of "1" representing "No Impact on a Hiring Decision", a rating of "2" representing "Minor Impact on a Hiring Decision", "3" representing "Moderate Impact on a Hiring Decision", and "4" representing "Significant Impact on a Hiring Decision."

Results
While 42 principals provided demographic data, 39 principals rated the first 40 statements of the questionnaire, and 38 rated the last 11.Data from principals providing demographic data only were removed from the data set, leaving 39 questionnaire responses for analysis.Mean ratings for each statement were calculated, and statements were then rank ordered from high to low.Mean ratings for principal responses ranged from 3.79 to 1.97, with only one mean rating falling below the 2.0 indicator (minor impact on hiring decision).Twenty-two items were rated in the 2.0 -2.99 range, between minor impact and moderate impact on hiring decision, and 28 items were rated in the 3.0-3.79range, moderate to significant impact on hiring decision.
Rank ordered items were divided into groupings representing half the distance between each Likert scale point to look for response patterns related to item characteristics.Twelve items received average ratings between 3.5 and 4.0; sixteen items were rated 3.0 -3.49; sixteen between 2.50 and 2.99; six were rated 2.0 -2.49 and one item was rated below 1.99.An initial attempt at categorizing the items within and across sets was made by looking for similarity of content of items within a given set.Factor analysis of responses was not an option because of the small sample size and number of items, so construct category verification was completed by asking four college colleagues to place each survey item into one of nine categories: Violence, Language, Racism, Pejorative to Persons with Disabilities, Pejorative to Homosexual Persons, Substance Use, Sexual Content, Too Much Information (TMI), and Other.The Other category was eventually eliminated because it was a minimally descriptive category and no item was rated as Other by all raters or a majority of the raters.Raters were asked to indicate the primary two categories into which each statement fit, and to provide category rankings for any items that could belong to more than one category.Four persons categorized each item into a maximum of two categories ranked for best fit, though most items (49) did not receive secondary category rankings from all four raters.There was 85.7% inter-rater agreement on the first choice construct category for 50 of the 51 items.Forty of these items were placed into the same first-choice category by all four raters (e.g., "suck a d---and call me in the morning" was placed into the Sexual Content category by four raters).There was only one item with little category agreement among the raters, the item ranked fiftieth in its likeliness to impact principals' hiring decisions (photograph description of Jim holding a shotgun and standing next to a stuffed and mounted squirrel).
An analysis of category co-occurrence was conducted for the eleven items that did not receive 100% first-choice category agreement by the four raters (seeTable 2).Six of these items were placed in the same first-choice category by three raters (e.g., "What u gonna do, n----" was placed in the Language category by three raters and Racism category by the fourth rater) and one item was placed into three categories, Racism, Too Much Information, and Other, by two raters each (photo of African-American female wearing a shirt that looks like a Wonder bread wrapper and caption: I'm Ms. Wonder bread!!!This was my halloween costume.Isn't it cute?).This item was, in the end, categorized in the TMI category since it did not clearly fit in any of the more descriptive categories.Four items were placed in two categories with equal frequency.Two of these received three ratings in the TMI category, and the more descriptive Language category, so they were ultimately placed in the Language category.The remaining two items were placed in two categories by three raters, and the researcher made the final categorization decision.Item 4, "Scuppernong and peach cider…" was placed in the Language category, with equal ratings received in the Sexual Content category, and Item 30, "big booty hoes…" was placed in the Sexual Content category, with equal ratings received in the Language category.Note: Numbers in cells indicate the number of raters who selected the corresponding category for the item.Blank cells reflect categories not chosen by any raters.Categories "violence" and "pejorative, homosexual" were not included in this table as they were not selected as areas of cooccurrence for any of these items.Asterisk* indicates category used for data analysis.
Two of the raters designated Racism as their first choice category for two of the 11 items without clear agreement among the raters, apparently indicating that Racism was not a strongly salient classification for the raters.Item 51, with two ratings in each of three categories, was placed in the TMI category, and Item 25 was placed in the Language category since it received three rater designations for this category and two for Racism.However, three other items that included racially pejorative language (the word "n----" or a variation of this word) were placed in the Language category (items 16 and 25) or Violence category (item 26) by the construct category raters.These three items received mean impact ratings (and rankings) of 3.45 ( 14), 3.59 (7) and 3.76 (2), all in the moderate to significant impact on hiring decision range.While these items could arguably have been placed in either the Racism category, or the category in which the researchers eventually placed them, it is clear that principals did perceive that racially pejorative language would have a moderate to significant impact on hiring decisions.
The category means for six of the eight construct categories was between 3.0-4.0,indicating a moderate to significant impact on hiring decision (see Table 3).The only category with a mean below this range was the TMI category.None of the 40 items for which there was first-choice category agreement for all four raters were designated as Racism.The four items in the Violence category were all ranked in the moderate to significant impact group, between 3.0 and 4.0.This was the sole category comprised entirely of items from the moderate to significant impact range.The Language items appeared in all rating groups; however, those appearing in the mild to moderate impact group included only mild language ("ur mama", "the finger" and "No farting") 2 .Stronger language items were rated in the moderate to significant impact group (e.g. , s---, n----, b----) 3 .Of the six categories with the highest mean item ranking, only the 2 Item 29: ur mama; Item 34: (image description) drawing of a smiley face giving the viewer "the finger"; Item 37: (image description) outline drawing of person standing with leg lifted encircled in red with diagonal line through picture and caption: No Farting 3 Item 2: "B----!I might be.";Item 3: "My true love is out there somewhere and they can go f---themselves"; Item 4: "Scuppernong and peach cider, Player #18, and vaginas"; Item 14: "Arms are for hugging.Boys are for kissing.Sluts are for dissing.And best friends are for when the boy is kissing the slut and all you really need is a hug"; Item 16 "Yo nyuggah sup witchu?";Item 21: "S---bullets.";Item 25: "What u gonna do, n----"; Item 31: "Celebrities walk on red carpet cuz they famous; I walk on toilet paper cuz I'm da s---." Pejorative to Homosexual Persons category, with two items, did not have an item ranked in the 3.5-4.0range.Items in the TMI category appeared only in the bottom Likert group categories; these items all had average impact ratings of 2.99 or below, placing them in the minor to no impact on hiring decision categories.From this analysis, then, it appears that not only does the type of content affect the impact on hiring decision rating, but the strength or severity of the content also plays a role.
Rating differences by sex and race of respondents were examined for each item using effect size (d), the standardized mean difference between groups (Cohen, 1998).Cohen's d is used to compare the means of two sets of data, and is reported in standard deviation units.Cohen offers the following guide for interpreting d as an effect size: d<0.3, small effect; d<0.5, medium effect; and d>0.8, large effect.For example, a value of d = 0.4 indicates that there is a moderate difference (0.4 standard deviation units) between the two groups.This measure is useful when one is attempting to determine if there are practical differences between two groups of data (Cohen, 1992).Summed across all items on the questionnaire used in the current study, the overall standardized mean difference between males and females was d = .32.This value indicates a small to moderate effect size, about a 1/3 standard deviation difference in mean ratings with females rating items as more likely to impact hiring decisions.Females rated 48 of the 51 items higher than males, and the remaining three items 4 received very similar mean ratings (largest d=.199 [Item 26: N----, I'll kill u.] with males rating this item as slightly more likely to impact hiring than females).African-American respondents generally rated the items as more likely to impact hiring decisions than did White respondents (d=.59; a moderate effect size) with five items 5 indicating little difference between the groups (largest d = .12[Item 19: happy halloween!!!This holiday rocks, I'm stoned, & I have candyyy], with White respondents rating this item as slightly more likely to impact hiring decisions than Black respondents).The two items most likely to impact hiring decisions, items 26 (N----, I'll kill u) and 28 (When all else fails, hit dat b----, kick dat ho, stomp that slut), were among those for both the sex and race analyses that showed least variation by group.That is, they were likely to impact the hiring decisions similarly for all groups, and received the two highest overall average ratings on likeliness to impact hiring decision.
In general, the stronger or more graphic statements were rated as more likely to influence hiring decisions than were the weaker statements (Table 4).For example in the violence category, "When all else fails, hit dat b----, kick dat ho, stomp dat slut" and "N----, I'll kill you" received higher ratings than other statements of violence such as "Im gon punch u in ur throat" and implied violence such as "[photo description] Jim holding shotgun and standing next to stuffed and mounted squirrel".Another example shows that some language "My true love is out there and they can go f---themselves" may be considered more objectionable (rank 3) than other language such as "s---bullets" (rank of 25).Pejorative language that related to people with disabilities was considered as less likely to have an influence on hiring decisions than statements related to violence or vulgar language, but more likely than statements considered pejorative to homosexual persons.The pejorative statement most likely to affect hiring decisions, pejorative to persons with disabilities, was in the first quartile, "U look flickted, retarted, and down syndromed" (first quartile) followed by an 4 Item 26: "N----, I'll kill u'; Item 28: "When all else fails, hit dat b----, kick dat ho, stomp that slut"; and Item 31: "Celebrities walk on red carpet cuz they famous, I walk on toilet paper cuz im da s---." 5 Item 1: "All we have left is what is ahead of us and we make our own destiny my love…so change what makes you unhappy.Smoke a cigarette, get drunk, read the newspaper, do whatever you do what makes you happy.";Item 6: "I'm feeling sexy.";Item 19: "happy halloween!!!This holiday rocks, I'm stoned & I have candyyy"; Item 26: "N----, I'll kill u."; Item 28: When all else fails, hit dat bitch, kick dat ho, stomp that slut."item pejorative to homosexual persons, "kid below me is gay.He's 'king of the circle jerk'" (second quartile).The remaining statements that were categorized as pejorative to persons with disabilities or homosexual persons all fell in the third quartile, are less "strong" and again, the disabilities statements are ranked as more likely to influence a hiring decision than the statements related to homosexuality.The statement "RETARD" was rated as more likely to have an impact on hiring decision (as were two other statements containing the word "retard") than was the statement "simply….thegayest thing ever" found in the third quartile.Construct: V = violence; L = language; X = sexual content; PH = pejorative to homosexuals; PD = pejorative to persons with disabilities; SU = substance use; R = racism; TMI = too much information

Discussion
The purpose of this study was to determine if principals thought their hiring decisions would be affected by information on teacher applicant SNS sites.To answer this question, principals were asked to rate the impact on hiring decision of a series of items found on pre-service teacher SNS.The items on the questionnaire included swear words, references to sexual content and alcohol, and comments that were pejorative to students with disabilities and to homosexual persons.
All items containing words considered to be swear or taboo words were found in the moderate to significant impact ranges of item rank with the exception of the item containing the word "farting."Of the ten most frequently used taboo words (Jay, 2009) five appeared in the survey items (f--k, s---, d---, a--, b----).In a 2008 investigation of the pragmatics of swearing, Jay and Janschewitz concluded that swearing is a complex act of communication that is influenced by physical and social setting, the topic of discussion, and the relationship between the speaker and listener.In particular, the authors note that "In formal contexts and with participants of unequal status, swearing is not expected" (p.285).It is likely, then, that school principals would not expect their employees nor prospective employees, teachers, to swear in the formal workplace context.While not all of the 51 statements on this instrument included swear or taboo words, many, if not all of them, included information that one would not expect applicants to share in a job interview situation (e.g., photos of hunting or of self in a bikini; comments about drinking alcoholic beverages).Furthermore, these topics and types of language may be considered, minimally, unprofessional and reflective of conduct unbecoming a teacher, and may even reflect dispositions that are undesirable in teachers.
Though dispositions represent attitudes and as such are difficult to assess directly (Henerson, Morris, Fitz-Gibbon, 1987;Oosterhof, 2010), education agencies are requiring colleges/schools of education to assess the professional dispositions of teacher candidates (CCSSO, p.31;GaPSC, 2009;NCATE, 2010).Another challenge in assessing dispositions is that there is not general agreement on the definition of the term (Singh & Stoloff, 2007;Stooksberry, Schussler, & Bercaw, 2009;Thornton, 2006;Welch et al., 2010), and from an assessment perspective, it is impossible to assess that which is not clearly defined (Welch et al., 2010).Furthermore, legal findings (Board of Education of the Warsaw Central School District, 1994;Dixon v. Clem, 2007;The Gallatin County Board of Education v. Mann, 1998) make it clear that teacher behavior does matter.Dispositions are typically assessed using behavioral indicators (NCATE, 2010;Welch et al., 2010) and the information found on SNS is just one behavioral indicator that might point to underlying professional dispositions and attitudes.Media reports indicate that employers are examining SNS as they make hiring decisions (Alfonsi, 2006;Amburgey, 2006;Bergstrom, 2008;Bock, 2008;Fuller, 2006;Jannsen, 2009).Teacher education faculty and administrators can better prepare teachers for the professional aspects of their careers when armed with data supporting the need for professional behavior in school and online.This is critical for inservice and preservice teachers since K-12 students are tech savvy and can access the SNS of teachers who are expected to uphold professional standards such as Standard 2 of the Georgia Code of Ethics for Educators: "An educator should always maintain a professional relationship with all students, both in and outside the classroom" and Standard 10: "An educator should demonstrate conduct that follows generally recognized professional standards.Unethical conduct is any conduct that impairs the certificate holder's ability to function professionally in his or her employment position or a pattern of behavior or conduct that is detrimental to the health, welfare, discipline, or morals of students" (Georgia Professional Standards Commission, 2009).Not only is it important for teachers to conduct themselves professionally in school, but this study also shows the importance of portraying a professional persona online.
As society engages in the widespread use of SNS that record and display indicators, such as comments or photos, of the professional conduct of our teachers, it is critical for school personnel and future teachers to consider this form of communication which is neither private nor fleeting as subject to the ethical and professional standards expected of professional educators.It is inevitable that some of this information will reflect behavior that is unbecoming to teachers and is also likely to undermine the functioning of the public employer, as established in Pickering (Pickering v. Board of Education, 1968).The data from this study will be useful to policy makers at the school, district, and state level as they consider policies related to teacher use of the internet and SNS.A recent search of 32 randomly selected Georgia district websites indicated that approximately one-third (10 of 32) of the districts had a posted policy related to teacher use of school internet facilities.While the policies related to district technology use ranged from extremely broad to very specific,6 none of the policies examined indicated any restrictions on personal use or information posted on personal SNS.Yet, personnel managers are examining SNS as they make hiring decisions (Alfonsi, 2006;Amburgey, 2006;Bergstrom, 2008;Bock, 2008;Fuller, 2006;Jannsen, 2009) and teachers are facing censure for material on their SNS (e. g., Associated Press, 2010;cbsatlanta.com, 2009;Geha & Sterling, 2009;Hui, 2010;Shapira, 2008).Lavelle (2010) suggests that every employer should develop a social media networking policy.Furthermore, states may need to get involved in helping to establish legislation that is consistent across districts for addressing concerns with teacher SNS and their First Amendment rights.Missouri is currently the only state with a mandated social networking policy (Estrada, 2010), though some districts have policies, such as the Hattiesburg, MS policy forbidding teachers to communicate with students via SNS (Associated press, 2008a).Estrada (2010) lays out a detailed recommendation for the state of California to develop a statute that would not only mandate privacy settings on teachers' personal SNS, but would also prohibit them from communicating with students via SNS, while still protecting their First Amendment rights.
While most states do not have social networking policies in place, it is also likely that not all universities have SNS use policies.Colleges of education, as well as school districts, would be wise to establish policies for SNS use by students before more lawsuits, such as that brought by Stacy Snyder against Millersville University, (Snyder v. Millersville University et al., 2007) are brought to contest disciplinary actions.For example, at GSU the Director of Field Experiences and Partnerships, in conjunction with the University Legal Affairs Office, crafted the following Statement on Internet BLOG Sites.
Participation in student teaching is a privilege and carries with it professional responsibilities.As a future professional educator, you are expected to maintain high standards of personal and professional ethics at all times and in all settings.Social networking sites, such as Facebook, MySpace, Xanga, and Friendster should represent your high personal and professional standards.If you have postings on any of these sites you must remember that they are public and may be viewed by school personnel and students.If school personnel find your postings to be unprofessional they can request your removal from their school.This request will be honored immediately and a new student teaching placement will not be identified until the following semester (2007).This statement is introduced in the orientation to the first field placement, the Pre-Professional Block (PPB) practicum, which takes place prior to the student's admission to the Teacher Education Program (TEP) and is included in the field placement information guidelines for every subsequent field experience.While faculty at GSU do not routinely monitor students' SNS, students are expected to maintain professional online personas and are dismissed from their practicum placement if objectionable material is reported to university personnel.Since students cannot be admitted to the TEP without successful completion of the PPB, conduct unbecoming a teacher as presented on a student's SNS could be instrumental in denying a student entrance to the TEP, if the student is unsuccessful in PPB because of SNS postings.
Demographic data from the current study indicate that the sample was consistent with the principal population of Georgia, though no data were gathered from principals in metropolitan areas.Therefore, while results may be generalized to principals throughout Georgia, there may be limited generalization to principals in metropolitan areas.This study does provide evidence that principals perceive that information on SNS would impact their hiring decisions, though describing how or if information on SNS impacts principal hiring decisions is beyond the scope of the current study.There is evidence that employers are using SNS to research job candidates (Grasz, 2009), however, this evidence comes from studies of corporate hiring practices indicating that the companies most likely to examine SNS are those related to information technology and professional and business services.Among the limitations of the current study, in addition to not determining if principals actually do use SNS information, is the decontextualized nature of the comments utilized on the survey.While the survey items did come from actual teacher education candidates' SNS, these were the only data presented to principals on the survey.It is possible that the statements would influence principals' hiring decisions differently than these results suggest if other employment data, such as resumes, interviews, character, etc., were part of the hiring decision data.It is also possible that a single inflammatory statement on a site with other neutral statements would have less perceived impact on a hiring decision than the series of decontextualized statements presented on this survey.One direction for future research is to provide principals or other hiring managers with a more complete employment dossier with SNS data embedded in the context of a resume and work samples and ask principals to rate the impact of the SNS statements in context.
While further study is needed to determine if principals actually do use information found on SNS in their hiring decisions, and if so, how this information is used, it is not unreasonable to believe that principals would be likely to use the information on SNS in much the same way that other hiring personnel do.Future work should also examine how education human resource administrators might utilize SNS data in their hiring decisions.The data from this study suggest that some information on SNS could influence a principal's hiring decision, and thus, that teachers and future teachers should consider keeping their SNS clean and professional.

Table 1 .
Respondent Demographic Data

Table 2 .
Areas of Co-occurrence for Eleven Items Not Rated in Same First Choice Category by Four Raters

Table 3 .
Construct Category Means and Representation in Likert Scale Grouping Sets Number of Items in Each Likert Scale Range Construct Category

Table 4
Survey Item Rank Ordering