USF Libraries
USF Digital Collections

The Pothole Lake fire


Material Information

The Pothole Lake fire an analysis of emotion in a successful emergency response
Series Title:
Quick response research report ;
Physical Description:
21, 8 p. : ; 28 cm.
Carley, Kathleen M
University of Colorado, Boulder -- Natural Hazards Research and Applications Information Center
University of Colorado, Boulder -- Institute of Behavioral Science
Natural Hazards Research and Applications Information Center, University of Colorado
Place of Publication:
Boulder, Colo
Publication Date:


Subjects / Keywords:
Wildfires -- Alaska -- Kenai Peninsula   ( lcsh )
Command and control systems -- Alaska -- Kenai Peninsula   ( lcsh )
Emergency management -- Psychological aspects   ( lcsh )
Fire fighters -- Psychology -- Alaska   ( lcsh )
Disaster victims -- Psychology -- Alaska   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Includes bibliographical references.
Additional Physical Form:
Also issued online as part of a joint project with the Louis de la Parte Florida Mental Health Institute (FMHI) Research Library's disaster mental health initiative.
Statement of Responsibility:
Kathleen Carley ... et al.
General Note:
Title from cover.
General Note:
"Institute of Behavioral Science #6."
General Note:
"January 1, 1994"--P. i.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001985085
oclc - 30804263
usfldc doi - F57-00021
usfldc handle - f57.21
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200361Ia 4500
controlfield tag 001 001985085
005 20090220161430.0
008 940721s1993 cou b 000 0 eng d
datafield ind1 8 ind2 024
subfield code a F57-00021
0 4 245
The Pothole Lake fire :
b an analysis of emotion in a successful emergency response /
Kathleen Carley ... [et al.]
Boulder, Colo. :
Natural Hazards Research and Applications Information Center, University of Colorado,
21, [8] p. ;
28 cm.
1 490
Quick response research report ;
v #64
Title from cover.
"Institute of Behavioral Science #6."
"January 1, 1994"--P. [i].
Includes bibliographical references.
Also issued online as part of a joint project with the Louis de la Parte Florida Mental Health Institute (FMHI) Research Library's disaster mental health initiative.
z Alaska
Kenai Peninsula.
Command and control systems
Kenai Peninsula.
Emergency management
x Psychological aspects.
Fire fighters
Disaster victims
Carley, Kathleen M.
2 710
University of Colorado, Boulder.
Natural Hazards Research and Applications Information Center.
University of Colorado, Boulder.
Institute of Behavioral Science.
t Natural Hazards Center Collection
Quick response research report ;
P68 1993


JJ NaturalHazards Research'andApplicationsInformationCenterCampusBox 482UniversityofColorado Boulder, Colorado 80309-0482HAzARD HOUSE (OPy The Pothole Lake Fire:AnAnalysis of EmotioninaSuccessful Emerqency Response By KathleenCarleyRuthCohnLaurieWaiselWilliamA.Wallace QUICK RESPONSE RESEARCHREPORT#64 J.993TheviewsexpressedinthisreportarethoseoftheauthorsandnotnecessarilythoseoftheNaturalHazardsCenterortheUniversityofColorado.InstituteofBehavioral Science#6 (303) Telefax: (303) 492-2151 BITNEr: HAZARDS@COLORADO Intemet: hazards@vaxf.Colorado.EDU


NaturalHazardsResearchandApplicationsInformationCenterCampusBox482UniversityofColoradoBoulder,Colorado80309,0482ThePotholeLakeFire:AnAnalysisofEmotioninaSuccessfulEmergencyResponseByKathleenCarleyRuthCohnLaurieWaiselWilliamA.Wallace QUICK RESPONSERESEARCHREPORT #64 1993.. rt f the Natural Hazards This publicatIon pa0I formation Center's ongoing Research&Applications Report Series. Quick Response Researc http://vW'N' TheviewsexpressedinthisreportarethoseoftheauthorsandnotnecessarilythoseoftheNaturalHazardsCenterortheUniversityofColorado.InstituteofBehavioralScience#6 (303) 492,6818 Telefax: (303) 492-2151 BITNElJ HAZARDS@COLORADO Internet: hazards@vaxf.Colorado.EDU


ABSTRACTThis study explores the emotional contentofan effective disaster response situation and its coordination by the Incident Command System. Twenty-seven firefighters and local residents were interviewed during a major wildfIre in 1991. The resulting videotapes were transcribed, coded, and analyzedfor their emotional content. Results indicated that a varietyoffactors led to the successofthe disaster response effort, including the useofthe Incident Command System, luck, the natureofthe disaster, and the cooperation of the town and local agencies. This methodologyisrecommended for future case studies researching emotions and crises. Keywords: Disaster Response, Emotions,Content Analysis, Incident Command System, Crisis ManagementINTRODUCTIONThis study is part ofanongoing researchproject (Carley, Harrald, and Wallace, 1993) exploring the relationship among infonnation technologies, emotion, and decision-making. Our objective in this portionofthe researchistoexplore the emotional content ofaneffective disaster response situation and its coordination by the Incident Command System. Our aim istoprovide a benchmark, and emotional profile,tobeused as a baseline for future analysesofthis type. One of the authors (Cohn) conducted twenty-seven videotaped interviews with frrefighting staff and local residents during the Pothole Lake Fire #103108inthe Kenai Peninsula near Cooper Landing, Alaska. Wewillanalyze the interviews using content analysis in combination with Heise's evaluation ratings for emotion-related words. Effective disaster response requires rapid decision-making under stress. Decision-making skills can be enhanced or hindered by human psychological responsestodisaster.Onthe one hand,anemergency situation can produce increased confidence in and commitmenttoa courseofaction; people "risetothe occasion."Onthe other hand, disasters can cause peopletohesitate ortobecome immobilized and can lessen creativity, especiallyintennsofconsidering potential solutions (Carley, Cohn, Harrald, and Wallace, 1993). This studyisunique in that it represents a codificationofinterview data collected during, not after,aneffective disaster response situation. The advantageofimmediate data collectionisthat written or verbal repons ofanevent that have been generated from memory are subjecttodistortion in terms of both facts and emotions. U sing a perfonnance evaluationofthercsteam, as well as residents' comments from the interviews,wewill demonstrate that this response effort was generally well-received and thoughttobesuccessful. Heise (1977) has noted that evaluation (good versus bad), potency (powerful versus powerless), and activity (active versus passive) are dimensions of affective response commontopeople around the even when they come from different cultures and speak different languages. He developed a dictionaryofaffect-related wordstowhichhewas able, through his research, to assign ratings ranging from-4to+4on eachofthe three dimensions.Inour1


research, we wanted to explore the rangeofpositivetonegative emotions present during the Pothole Lake Fire; therefore, we used ratings from Heise's evaluation dimension to extract emotional content from text.THE POTHOLE LAKE FIREFireSuppressionin Alaska Because Alaska contains large areasofundeveloped, sparsely populated land that is relatively inaccessible, the state relies on a decade-old master flIC management plantodetennine appropriate response to fires (Enders, 1991). Naturally occurring wildfires often are beneficial, clearing the way for new growthofgrasses and shrubs, thus providing new food sources for animals such as foxes, birds, and moose (Reeves, 1991). Areas are designated for full, modified, or limited fire suppression, depending on how desirableorundesirable fire inthose locationswould be (Enders, 1991). Areas designated for limited fire suppression are locations where the benefitsoffire outweigh risks to resources. Fires often help restore the healthofailing forests; when fires are suppressed, the forests age and become susceptibletoinsect attacks (Kizzia, 1991). Three agencies fight fire in Alaska: the Alaska Fire Service, an agencyofthe federal BureauofLand Management; the StateofAlaska DivisionofForestry (DOF); and the United States Forest Service. Fires occurring along the limited road networks usually are handled by the DOF.TheAlaska Fire Service manages fire suppression in Alaska's less accessible areas. Backup crews are called in fromthe Lower 48 when necessary (Enders, 1991). Allofthese agencies have selected and adopted the Incident Command System as a meansoforganization for fighting fire (Lien, 1993). All personnel who work for these agencies are trained in ICS firefighting methods and know when an ICS team needstobe mobilized. In ordertocall in an ICS team, authorities contact the DOF in Anchorage, which in turn call the Alaska Interagency Fire Coordination Center (AlFCC) to request the typeofteam needed. The AIFCC has three established Type II teams (teams trained to respondtomoderately severe fires) on a two-hour call. All personnel on these teams work for either theUSForest Serviceorthe StateofAlaska DivisionofNatural Resources. Wages for a firefighter's first eight hours on an incident are paid by that firefighter's employer. After that, all wages and expenses are paid by the ownerofthe land on which the fire is being fought.Inthe caseofthe Pothole Lake Fire, the expense was borne by the United States ForestService (Lien, 1993).DescriptionofthePotholeLakeFireThePothole Lake Fire was first reported on May 19, 1991 as covering about ten acres near Pothole Lake in the Kenai Peninsula in southern Alaska. Because the flIC started in a limited suppression zone, it was monitored and allowedtoburn freely for the first few days (Kizzia, 1991). The fire, which authorities think may have been started by local bear hunters (Enders, 1991), fed on spruce trees that had been killed by beetles. Several days later, the fire had 2


grown to fifty acres and then suddenly balloonedtoeight hundred, threatening the town of Cooper Landing and the Kenai and Russian River fishing grounds. By May 24, the fire was declared to be at the Class II level, and its size was estimated at 3600 acres. On May 27, when the fire was mapped at 8400 acres(laterremappedto7900 acres), aaassI team, which consistsofthe most experiencedfiresuppression staff, was called in. Campgrounds and highways were evacuated, and public meetings were held daily for residents of the townofCooper Landing (Cohn andWallace, 1991). Firefighting crews workedtobuild a line around the perimeterofthe fire. A wind shift, cooler temperatures, and higher humidity assisted the firefighting effort. On JuneI,the fire was declared contained, and on June 3, the fire was declared controlled. On June 4, the fire was reducedtoaassilland then was turned over to local fire authoritiesonJune 23. Isolated hot spots continuedtobemonitored for several months, until the fire was declared completely out on October 31, 1991 (Frre Narrative, 1991). 650 personnel from thiny crews had fought the fire (Frre Overhead Perfonnance Rating, 1991).FIELD RESEARCHThe fire-related personnel and activities were observed by Cohn from May 30,1991to June 4, 1991, while the Pothole Lake fire wasinprogress. She had access toallfIrefighting and support personnel and videotaped interviews with fIrefighters, support staff, local residents, and land managers,aslistedinTable1.In-depth interviews were conducted using non-directive open-ended questionsina flexible fonnat to gather descriptive datainthe interviewee's own words. Sample questions included:"What'syourname?""Whatis a smoke jumper?" "Anything particularly unique about this situation here?" "What kindofconcerns does a populated area present for a fIre management?""What'sthe experience been for you and the crewonthis particular incident?" "Have you done this for a long time?" "Are there any unusual circumstances with this particular incident?" "What does resource management mean?" "Is this your first fIre inAlaska?" "How.hasitbeen working with these interagency teams from different organizations?" "After this, what will youbedoing this summer?""How'dyou hear about this fIre?" "Howdid the crew handle that kind of situation?" Sometimes reflective comments were madetoencourage the intervieweetocontinue talking: "Moving camp cenainly sounds like arealchallenge.""You'vedone a lot of travel." "So you were literally jumping in the fire."3


"Soundslike that was quite an experience for the crew."Theresponses obtained varied depending on the interviewee's verbal ability, willingnesstobe andcomfonlevel with being videotaped.Weexpected to find a significant amountofvisible distress surrounding such a threatening incident and was surprisedtohear most interviewees express satisfaction with the progressoffire suppression activities.Weattribute this initial finding to thestaffshoursoftraining,totheir yearsofexperience, andtothe smooth operationofthe Incident Command System.Theinterviews averaged 19 minutes in length and ranged from one minuteto84 minutes.The8-millimeter videotape was transcribed to diskette with one file for each interview. Various lists, plans, directories, maps, and newspaper articles from theAnchorageDaily News and theAnchorageTimeswere also collected.3THEINCIDENT COMMAND SYSTEM: STRUCTURE AND PHILOSOPHYTheIncident Command System (ICS), originally developed in responsetolarge firesinSouthern California in 1970 (Hazardous Materials Operations for First Responders, 1990),isa national paramilitary emergency response network that can be mobilized to manage emergenciesofany type. The system consistsofa groupofproceduresfor managing personnel, equipment, facilities, and communications during an emergency (New Jersey Hazmat Emergency Response Course, 1990). It is military in its tightly hierarchical structure and minimal spanofcontrol, while the unified command structureoflocal agency officials and residents is distinctly non-military, as is its completely voluntary aspect: anyone can quit at any time.Itis importanttoemphasize that ICS is a response system, a protocol, rather than an existing ongoing organization. Administratively, ICS ispanofthe National Inter-Agency Incident Management System (NIIMS), a branchofthe Federal Emergency Management Association (FEMA). NIIMS is an umbrella organization intendedtoprovide a common system to be used by risk management agencies (including, but not limited to, flre protection organizations) at all levels.3 These d:uaareavailable fromProfessorW.A.W Decision Sciences and Enainccrina Systems. Rensselaer PolytechnicInstilUte. Troy.NY. 12180-3590. 4


Within NIIMS are the National Inter-Agency Frre Qualifications system (NIFQS) and ICS. NIFQS develops training courses for positions within ICS as well as standards for certification and qualification.lCS,built cooperativelybylocal, state, and federal agencies,isbased on thestructureofa large fire organization, as developed by Federal frre protection agencies. Unlike most fire organizations, however, ICS is designed to be used for different kindsofemergencies, and for incidentsofvarying sizes. Local emergency management personnel have plans in place to mobilize teams as needed for emergenciesofdifferent kind and scope.Teamsare classified according to the typeofemergency they are qualifiedtorespondto.Federallawrequires thatIeSbe used to manage hazardous materials incidents (Hazardous Materials Operations for First Responders, 1990).Atthe present time, Alaska has only frrefightingIeSteams, although the BureauofLand Management is working on developing a cross-trained all-risk team.TheUnited States Park Service has a standing all-risk team (Lien, 1993).Thebasic philosophyofIeSaddresses scope, organizational structure, and operating principles. ICS can operate in a single location or over multiple jurisdictions, with one agencyormany, andmustbe able to function in any areaofthe country. Its organizational structure must be adaptable to emergencies caused by "fires, floods, earthquakes, hurricanes, tornados, tidal waves, riots, spillsofhazardous materials, and other naturalorman-caused incidents" (The Incident Command System and Structural Firefighting, 1984, B-6.) .. Its hierarchical structure must be able to expand and contract effectively as required by the changing scopeoftheincidentICS should be able to adapttonew technology and should be internally consistent in regard to organization, tenmnology, and procedures. OneofJCS's top prioritiesisto cause as little disruption as possibletothe communities and systems within which they work. Finally, JCS must be able to meet allofthe above requirements while maintaininglowoperational expenses.TheICS structure for any given incident is unique to that incident and is constructed in amodularfashion.TheIncident Commander (IC) builds the organization's staff from the top down.TheIeappoints Section Chiefs who become responsible for all further developmentoftheir section. Sections which may be created as needed are Operations, Planning, Logistics, and Finance. Each section may have up to three sub-levels: Branch, Division, andUnitAn Incident Command team canbestaffed by personnel from a varietyofemergency services and agencies: firefighters, Red Cross personnel, medical personnel, local emergencymanagementofficers,orhazardous materials experts. Integrated communications are managed by the Communications Unit in the Service Branchofthe Logistics Section; media may include radio networks, on-site telephone, public address, and off-site systems. Separate radio networks are organized by function for the following areas:command,tactical, support, ground to air, and air to air. All communications are in plain English, and no codes are used.Whenan incident occurs, and local emergency management resources are not adequate, one 5


or more agencies may callinan ICS team. Incidents that cross jurisdictional boundaries or involve multiple agencies require a unified command structure, in which all agencies involved participate in determining the overall objectives in managing the incident. Together, they select strategies and plan tactical operations in such a way astouse resources with maximum efficiency. The unified command structure usually consistsofkey officials from each jurisdiction or department heads from different areas within a single jurisdiction, anditmay include local landowners as well. Although the unified command structure as a group is responsible for planning, the implementationofthe plan is directed by the Operations Section Chief, who generally comes from the agency with the greatest jurisdictional involvement and who reports to the Incident Commander. Planning by group followed by implementation under the authorityofa single individual is an element essentialtoICS structure. When an ICS team is called intosuppress a fire, the fire and the team are classified as Type I, Type II, or Type III, with I being least severe andillbeing most severe. In placeof"Type," the words "Level" or "Class" are sometimes used. Most people on the teams work at other jobs when they are not respondingtoan ICS incident, many. though not all, in the fieldoffrrefighting. Praise for the efficiency and effectivenessofthe Incident Command System has led FEMAtoexplore using the systeminpermanent organizations that handle routine situations, accordingtoDave Liebersbach (1993). He emphasized that the needsofanorganization are differentinroutine and non-routine situations, citing minimal spanofcontrol asanexampleofa principle that would not be appropriate for day-to-day operations. He pointed out that ICS can, however, be effective for non-routine events that are not emergencies; ICS handled the Pearl 'Harbor 50th Anniversary celebration, which lasted six months and ranged in location from Washington, D.C.toGuam.CONTENT ANALYSISWeuse both quantitative and qualitative content analysistoexplore the emotional contentofthe interview text. Accordingtoone definition, "Content analysis is any research technique for making inferences by systematically and objectively identifying specified characteristics within text" (Stone, Dunphy, Smith, and Ogilvie, 1966,p.5). This definition,ifeditedtoexclude the word "objectively," would applytoqualitative as wellastoquantitative content analysis. The inferenceswewishtomake concern the emotional atmosphere in an effective disaster response effort suchasthe Pothole Lake Fire. The specified characteristics within the textthat we wantedtoidentify are emotion-related words and concepts. For the quantitative analysis, these characteristics were systematically and objectivelyidentified-6


through the useofa software program that does word frequency counts.Thelistofemotion related words tobesearched for was based on Heise's research. For the qualitative analysis, emotion-related concepts were systematically identified by studying the text and extracting relevant passages. Content analysismaybeimplicitorexplicit. Explicit analysis searches for wordsorphrases which actually occur in the text, while implicit analysis identifies concepts which occur in the text only by implication (Kang, Wallace, and Carley, 1993). For example, "John loves Mary" contains the explicit emotion "love," while "John beat Steve" implies that Steve and John may be adversaries. Explicit analysis is easier to automate and quantify, but may miss infonnation thatisimplied rather than stated.Wewill use both approaches: explicit analysis with word frequency counts and implicit analysis through studyofthe text.Themain advantagesofusing computer-aided word frequency counts are speed and the ability to handle large volumesofdata. The primary disadvantageofword frequency counts is that they cannot distinguish between homographs, words that are spelled the same but have different meanings when used in different contexts. Some content analysis software packages overcome this problem by using an indexing feature which lists sentences that contain words whose meaning depends on their context (Kanget a/.,1993). The researcher then reviews the questionable sentences and makes a decision about them. This kindofprogram was not used in this research; however, the problemofhomographs is addressed later in this paper. Another disadvantage is that the word list is subjective on the partofthe researcher.Wehave -attempted to reduce this subjectivity by using an emotion-word list developed by Heise as a resultofhis research in affect control theory. In order to conduct a frequency count with the program used in this analysis, two input files are required: a concept list and a text list. The text list is simply a listoftext file names to be searched. In this study, each transcribed interview has its own text me. A concept is a single idea that may be represented by a single word or by a phrase concatenated by a "_" symbol. Each concept consistsofa base "word," which is a word or phrase that represents a key concept, and discrete "words," which represent synonymous wordsorphrases equivalent to the base word. For example, the base word or concept "talk_to" is associated with the discrete words "talked_to," "talkinLto," and "talks_to." Although the program perfonns primarily explicit analysis, the analysis also is implicit as a resultofthe structureofthe concept list, which links base concepts with related discrete words. When the program finds oneofthe discrete words, the base concepttowhich that word is connected is coded, and thus a concept has been found through implication, rather than through explicit use. The program can handle subtletiesofspacing and punctuation and does not match to partial word strings, thus avoiding inaccurate matches based on embedded words. For example, "fear monger" would not be identified as an instanceof"fear".Theprogram begins by counting the numberofoccurrencesofconcepts in the texts and produces an output file containing all base and discrete words along with their frequency counts. The number beside each discrete word is a frequency count for that word alone, 7


while the number beside each base word is thesumofall counts for its associated discrete words. Since each base word is also included as oneofits discrete words. it is clear how many occurrencesofthe concept come from the actual base word as well as from the related discrete words.Theprogram then strips from the output file any baseordiscrete word that did notoccurinanyofthe texts being processed, resulting in a reduced word list and frequency count. Additional output files provide a listofwords removedand basic statistical data about the concepts found.Theoutput me "wordf.out" consistsofa listofallbase and discrete words listed alphabetically along with their respective counts. There is one word for each base word and discrete wordandone column for each text me orinterview. A final column identifies base words with a"I"anddiscrete words with a "0". Thecounts for base words are the sumsofthe countsofall associated discrete words.Theoutput file wordf.out was used as the basis for further statistical analysis.QuantitativeAnalvsisThefile wordf.out was reduced to a listofall the base words. each with the total numberoftimes each base word and its related discrete words were used. Heise's evaluation ratings were assigned to each word. There were 275 words in the output me that corresponded exactly to words on Heise'slistAnother twenty-five words in the output file were similar enough to words on the Heise list that the ratings for those words were used. Thirteen words in the output file had no matchjudgedclose enough on the Heise list and therefore were discarded from the analysis.Theevaluation rating for each word was multiplied by the numberoftimes each interviewee used the word. These products were then totalled for each interviewee. providing a raw E-score. and divided by the total word count for each interview to yield an adjusted E-score. Content analysis as a methodology has a numberofdata-related problems that must be resolved.Thebiggest problem in this study is the inabilityofthe software to distinguish between useofwords in different contexts. This may include words used as proper names as well as words that have more than one meaning in common usage. An exampleofthe former problem would be picking up the phrase "Incident Command System" as an instanceofthe word"command."Since"command"has a Heise E-ratingof-.32 for males and -.35 for females. frequent useofthe phrase "Incident Command System" could negatively skew the results. Fortunately, this turned out nottobea problem. since the interviewees tendedtouse the abbreviation"ICS"rather than the full name.Themultiple meaning problem arises for common words. such as"like"or"kind," and also for less common words that are used unusually often due to the natureoftheenvironmentWords in the latter category might include "command" (as separate from "Incident Command System"). "control," "direct,"and"out"Since this research involves a paramilitary organization engaged in frrefighting, these words were expected to occur more frequently than usual and thus had the potentialofskewing the E-scores. In fact, the data showed that the words"command,""control," and "direct" did have a fairly high incidence; however, it was not possible from examining the texttostate positively that these words had no emotional8


connotation, and therefore they were not removed from the analysis. The word "out" was the most frequently used, with a countof404, but since Heise had no E-rating for this word, the issueofinclusion wasmootThe words "like" (used 389 times) and "kind" (used138times) also created a multiple meaning problem. For bothofthese words, we examined individual instances in the texttodetermine the best waytohandle them. We found that "like" was used both as a colloquialism ("Like, we ate dinner'') and asanemotion-related word ("I like my job''). For this reason, "like" was left in the analysis. "Kind," however, was removed, since nearly all instancesofits use were colloquial ("kindof'), notanemotion-related usage. The same problem occurred with the word "mean," which was used 76 times. All usesofthis word were in the contextof"Imean," rather than in the word's emotional sense. Therefore, "mean" also was removed from the analysis. The interviewer'S words were stripped from the texts prior to the runningofthe content analysis program. Basic statistics on word frequency usage are displayed in Table 2 and Table3.The words "out" and "want" were two of the thirteen words with no suitable match in Heise's dictionary; therefore, they have no E-rating. Because each adjusted E-score is based on 313 individual word scores, the Central Limit Theorem applies, and the adjusted E-scores canbeassumed tobeapproximately normally distributed. Table 4 lists the adjusted E-scores for each interviewee. Twenty-six outofthe twenty-seven scores are positive, implyinganoverall presenceofpositive emotion, which also was the impression we got from studying the text; however, since this was the first projectofthis sort, we havenoprevious E-score profiles with whichtocompare our results. Although Maronbanks had a negative adjusted E-score, his interview does not seem negative in tone. Because we had no similar data from other incidents with whichtocompare these results, we performed an internal statistical analysis.9


Weclassified each interviewee accordingtothree variables: gender, role, and position. The sub-classes within gender were male and female; the sub-classes within role were resident and staff; and the sub-classes within position were managerial/supervisory and non managerial/supervisory. Each interviewee was assignedtoone subgroup in eachofthe three classifications. The interview called Marty was removed from this sectionofthe analysis, because there actually were two interviewees, both named Many, oneofwhom held a supervisory position and the otherofwhom did not. Dorothy Todd is a resident who worked in planning support, thus technically placing her in both the staff and resident categories. Reviewing her interview, however, we found that she spoke from the perspectiveofa resident and therefore put her in that category. Bruce VanZee wasina similar situation; but his interview dealt almost exclusively with his position as Chugach National Forest Manager, and so he was classified as staff. The assignments are listed in Table1.Statistical Analysis Software was usedtotest for significant effectsofthe three classifications. Table 5 lists hypotheses and results.Wetested for main effects of each of the three variables with hypotheses that compared the meansofthe two sub-groups within each classification variable. We tested also for ,interaction effects among all three variables. The results indicated a main effect for gender,andnomain effects for role or position, nor any interaction effects. Testingonrole and ,:'position alone yieldednomain effects. While there was a strong indicationofa main effect for gender, our inferences were limited by what the sample represents; six women and15men were interviewed. We were doing a retroactive analysisofdata, and the interviewees were not picked with this particular kind of analysis in mind.Ifthis research weretobe repeated using the same classifications,werecommend selectinganequal numberofwomen and men, using a larger sample size, and choosing a demographic cross-sectionofparticipants. The analysis yielded univariate statistics for each cellasshown in Table6.The areas for male and female resident managers are blocked out, because only staff held managerial or supervisory positions. The cell for the standard deviation for female staff managers is blocked out because thereisonly one occupant of that cell,sothereisnobasis for comparison. Also, main effects could not be compared for that cell, for the same reason.10


QualitativeAnalysis After studying the transcribed interviews, we selected nine broad emotion-related themes for further analysis.Theareas, along with the numbersofinterviewees whotalkedabout them, arelisted in Table 7. A feelingofpride, pleasure, and/or excitement about their jobs, or about the Pothole Lake fIre, was expressed by twelve interviewees, often in strong terms. Some sample comments include: Tupper: Anderson: Knowles: Chittendon: Hasselquist: Hasselquist: Marty: bestjobontheplanet""I'vereally enjoyed's beenanexcellentexperience for me..""rwent to the town meeting here the other night and for meitwas a pleasure. It was too bad the fIrefIghters out there can'tgotothose...then the thankyous come along and the fIrefIghters never really gettoseethat""It's gratifying and it's challenging and I likeit""Youhavetopass the PT [physical training] testtoeven be abletomake it on the crew. So there's a lotofpride in that and being really..prepared and the standards the crew boss sets make it more organized and more structured. Dave [Jandt] had real high standards.""Ilove it...I'm so glad I madethis decision... It's alwaysanadventure."'s likeanadrenaline rush...You know and you're out in the fIre and it's burning and it's hot, it's going and everybody's pumpedup..." Many people expressed a levelofenthusiasm approaching zeal about their jobs. Several proudly recounted detailsofdangerous experiences or demanding training they'd been through. Often, they soundasthough they thrive on the physical risks involved. There have always been people who gravitatetoprofessions with an elementofdanger, people who fInd risk, physical challenge, and intense experiences invigorating and exciting. Jandt talked about the "surreal" and "quasi-religious" aspects of a cenain dangerous situation. He described a "real heightened senseofawarenessofthe surroundings." It is common knowledge that.11


dangerous situations can produce a naturally-occurring "high," as Marty (see above comments) described. For people who enjoy this levelofintensity, the heightened awareness it produces (see Jandt above) may enhance theirjobperformance. This could be one reason why this frrefighting effon went well: these firefighters clearly love their jobs and are proudofthe demanding work they are abletodo. Several people expressed pleasure specifically about the experienceofworking with others. Thereissome overlap here with the topic of group cohesion, and these comments express both positive emotions and also pleasure in contact with other people. These comments included: D. Todd: Perry: Stordal: "These people are really special." "I think working interagency is a real rewarding experience." "It's been really nice. They're a really good groupofpeople." It seems that working together with other people is a major sourceofpositive emotion for both the firefighters and the residents. Robertson talked about how nice it wastowork with people that he already knew:"Idon't havetogothrough the mating game with them to know personalities, introduce myself. They know how I think. 1 know how they think. We can just jumpin."Some people used the metaphoroffamily: Marty:'s like having brothers and sisters, twentyoftheminall it's like for some reason there's a bonding...everybody's unified together Hasselquist:"Theyreally are like brothers. They're real special." Perry talked about the primitive facilities and simple meals and his sense that these hardships enhanced a senseoftogetherness. Several people mentioned that the conditionsofthis fire were more primitive than the conditionsofmost Lower 48 fires. Hasselquist said, "Down south in the Lower 48 it's a lot more clustered and civilized. They have caterers. They have showers...Up here...we carry our own tents, sleeping bag, MRE's -we'll eat those for three days and then they usually fly us some fresh food. You do your own cooking. You setupyour own camp. You don't have all that other [stuff]...and we seemtoget a lot more work done." Group living under primitive conditions is knowntobuild group cohesion; that is the basic premiseofthe Outward Bound program, a program for adolescents designedtobuild trust and leadership skills. Group cohesion cenainly isanessential partofany firefighting effort, and Hasselquist made a direct connection between the primitive conditions and her sense that somehow, the hardships help them do their jobs better.Ifthis chainofcausality holds (primitive conditions contributetogroup cohesion, which improvesjobperformance), then this may help explain the high effectivenessofresponsetothis particular fire. 12


Other comments relating to group cohesion include: Hasselquist: ...the crew was madetoall interweave, you know. Andifone person is missing, you really notice it...that bonding, you know, cause you're all so interdependent on each other. And it can runrealsmooth." Jandt: "Just even after that day [of a dangerous situation] I saw a lotofthe camaraderie come together." Jandt went ontosay that the lead foreman, in handling this particular situation, acted instinctively as a resultofhis training and experience, but that he, Jandt, believed that the foreman's actions resulted also from the strong relationship among allofthe crew members. Stordal: "It's been really nice. They're a really good groupofpeople. Real helpful and they're really into doing this. They get excited when fire season comes along, they're, Oh, let's go put out this fire." On the more general topicofgroup dynamics, Jandt commented: "I thinkifpeople are happy doing what they're doing, they're goingtoproduce a lotofgood work. And, I think as a supervisor,tobeabletoanticipate the feelings the reactions, the developments thatgoonwithin a crew is necessary." Several people talked about how good they felttobe helping other people or the land: Stordal: Barnet: Hasselquist: "I'm glad we were abletodo something. Do our partinsaving thepark."'s a beautiful area and I'm gladwewere abletohelp them save some of the resources that were here, that were at risk." "I'm not just hereasa tourist I'm heretoprotect a village." People often derive emotional satisfaction from being of servicetoothers. Many professions offer different degreesofopportunitiestobeofassistancetopeople, places, or animals. Firefighters help in oneofthe most fundamental ways possible: they save lives, homes, and land. Their well-justified pride in their profession may leadtoa commitmenttodoing the bestjobpossible. Liebersbach expressedanattentivenesstothe needs of the town: ...the success may havetobemeasured in something other than just putting out the fire...itis threatening the community but at the same time, probablyasimportanttothem is that we don't damage he community with the people we have or equipment, our vehicles speeding through the town tryingtogettothe frre runs that's a pretty serious situation and probably much more serious than the fire itself." ... pan ofour objectives...weretoget this done with minimum impact on the town, which we did...the last stageofany rescue istoshoot the rescuer and basically this fire is done here and the biggest concern I have right nowisnot the fire but it's the impactonthe13


communityofCooper Landing and we're tryingtoget ourselves out of here. .we need to be outofthe way [for the upcoming tourist season]." OneofICS's operational priorities istoavoid disruption to communities andtobeattentive to local needs. As Incident Commander, Liebersbach was responsible for the overall operation, and his emphasison this aspect illustrates the useofICS principles in the field. The non disruption policy probably contributed strongly to acceptanceofand cooperation with ICS by local residents. These people already were threatened by theintrusionofthe fIre into their lives; any organization that wants tobesuccessful not only in fighting the fire, but also in helpingtoalleviate the distressofthe people needstoanendtohuman needs as well as to frrefighting needs. Such dedication and enthusiasm are not possible without strong support from the sponsoring organization. Four interviewees mentioned their senseoffeeling taken careofby ICS in general, and by their colleagues and supervisors in particular. Some comments in this area include: Klein: Perry: Jandt: Hasselquist: "They've taken good careofus outhere." ...the overhead team person...cannot forget the needs of the don't ask anythingofyour subordinates that you would not do yourself. And I think that's pretty good management philosophy andrdo that not only on fIres but I also do that in my everyday job." ...our niche in this organization istogoout and do the hands-on work...somebody has food here for us..Somebody has water. Somebody...tells us wheretocamp. Somebody picksupour garbage." "Well we have really good bosses. They knew what was going on. They kept people from panicking...[The crew boss] gave everyone that night to just think about it and the next day he talkedtous about [the close call]." These comments demonstrate different kindsof"taking careof'thatgoon withinrcs.Perry, speaking as a Resource Unit Leader, was well awareofhis responsibility to the people he commanded. Jandt was saying that because logistics were taken care of, the firefighters were able to concentrateontheir jobs. Although he was a division supervisor, he was referring nottohow he took careofhis people, buttohow the ICS organization took careofeveryone by clearly identifying scopesofresponsibility and making sure that all important tasks are assignedtosomeone. Hasselquist was referringtoher personal experience with an unusually dangerous situation; her comment illustrates that caretaking occursonan emotional levelaswell, and that supervisors were aware ofand concerned about people's feelings. In additiontothe feeling of being properly cared for, another element essentialtoa disaster relief effort is good communications. No fewer than seven interviewees discussed this topic, albeit ofteninresponsetoa communications-related question from the interviewer. Liebersbach: "...before we even got here, they already had a town meeting...going so 14


that we could meet with the town people, disseminate information to them about what was occurring, receive from them their concerns, their worries and whatnot so information was going on." Phil Todd: "They did an excellent keeping the community informed...They made an excellent effortatdoingthalEven though there wasn't a lotofpeople that went to all the meetings, there was people that did go that then spread the word to the community...and to me it was a comforting thing every nighttogo over andtolistentotheir updates. It only took twenty minutes or so, buttomeit was a very comforting thing." Painter:"Theyhad meetingsatthe school every night..You could come and ask them questions any time, which I did and...somebody was always glad to talk to you and they...had several phone lines put in [our library] and published phone numbers in thenewspapers so people could call in at anytime:'Ground: "...they did a very good jobofkeeping us [informed] where the fIre was, where they were working, what they were trying to do, when they would get it under control, when they would get it contained..." Hasselquist:"Backtothe communications thing and this whole organization, I think this fire's really been pretty well organized and communicated." Other interviewees who discussed communication were Dave Jandt and Lindsey Lien. Lien talked about hisjobas communications director and Jandt emphasized the imponanceofgood communication, "especially when there's a long chain of command." Liebersbach's comments illuminate an important but sometimes ignored facetofthe process: good communications require effort on both sides. The people of Cooper Landing tookanactive part in receiving'theirinformation; they already had a process in place when thercsClass I team arrived. This allowed thercspeople to concentrate on collecting and disseminating the information, rather than on settingupthe meetings. The comments of Todd, Painter, and Ground gave the residents' perspective on communications; clearly, they were very pleased. Painter talked about the accessibilityofinformation, and Phil Todd spoke about the emotional importanceofbeing informed. Ground's comment shows that ICS was abletoclearly communicate technicalflIefighting informationtolay people, a task that can be difficult in the midstof tunnoil andupselOneofthe ways that ICS takes care of its staff is by offering high-quality training. Jandt commented that "...[this has beena]real good fire in terms of training." His crew had been well-trained, and he saw the benefitofthat training during the Pothole Lake Fire. Hasselquist, who was in Jandt's crew, said, "This is like special training and I think there's only like two crews in the country that are getting extra, extra special flIe training. And we are one." Hasselquist's pleasure in the training also contains a good measure of pride. Perry talked about his workasa learning experience, rather than about specific training: "'s good because you teach and you learn at the same time andrbring someofmy experiences from the Lower48toAlaska and I intendontaking15


someoftheir management principles back to the Lower 48 when Igo."Chittendon, Logistics Section Chief commented that,"Ifyou have the staffing and they're experienced, it should run as smooth as possible, you know. Like I said, you're always going to have snafus and stuff like that but uh, the organization's there." Chittendon relies on experience and ICS training and structuretonavigate the unpredictabilitiesoffire suppression. Such highly stressful work performed with such pride inevitably leads to a certain amountofcompetition and even a senseofelitism. Tupper, a Smoke Jumper, compared his group to the Hotshots (the group to which Hasselquist belonged): "It's easiertohandle Hotshots than it is to handle Smoke Jumpers..." He was proudofthe relative independenceofhis group, which was n unlike the Hotshot crews...[who] have definitive leaders who are one guy's in charge period." Amatin, a local frrefighter, seemed to be speaking for his colleagues when he said, "...once in a while they always say Hotshot crews are better than us, but once in a while I don't think so...they go to training for it butitseems like they've been trainingalltheirlives."There was a senseofenvy here, perhaps for the professional development that is provided by ICS, as well as the feeling that "they've been training, but we've beendoing.".Anarticle inTheAnchorageDaily News stated, "Idle firefighters here grumbled last week at the news that 80oftheir competitors from Outside were hiredtohelp put out the Pothole Fire on the Kenai Peninsula," (Spencer, 1991). The fire "fueled a shon-term mini-economic boom in Cooper Landing...About $45,000 has gone for local firefighters' wages and fire depanment equipment," (Ellis, 1991). It is understandable that local frrefighters not called in would resent the work being given to outsiders. Given the threatofthe frre behavior, the TypeITIncident Commander, Ed Edmondson, made the decisiontobringinadditional specially trained Type I crews. Alaska has only two Type I crews, and bothofthese already were working on the Pothole Lake frre. Feelings over the locals' being left out seemed not, however, to impinge negatively on the general perceptionofhow the frre was handled. Liebersbach, the TypeillIncident Commander, cenainly was sensitivetolocal issues and did not want to repeat mistakes make in other disaster response situations: "The overlooked Valdez was the impact on town and the people there...they made a lotofmoney, but their whole town was taken away from them by the responders to theincident"Another theme that emerged from the interviews was that this fire was beneficialtotheforestWells:"Ithink it's a very beneficial fire...when the forest getstothe point that it is so old andsothick and tangled and everything, something hastobe done because when it gets that bad, there's nothing that can live there. I mean there's a few birds, but even mostofthe birds prefer a younger forest that's not quite16


so thick. They enjoy the heavy covertoproteCt them from their predators and stuff and when it gets real thick it gets too hard for even themtoget around in. So fire is good in some circumstances..that canbecontrolled." Doshier, the Managerofthe Kenai National Wildlife Refuge, "We use fIre as a management tool.from a resource standpoint, itwas doing good things. It's taking out the old, mature, dead forest, and, you know. new growth is more beneficial. It's more nutritious, and it's better for wildlife. You need the variety. You don't needallofone kindofforest. So breaking itupwith burned area is good..There's no doubt about it. It's been's a good fire from a wildlife standpoint. The bottom line is fire for the Refuge is good." VanZee. Managerofthe Chugach National Forest, had much the same perspective: is a natural process and the burn suchasthey have will be of benefit to the Refuge. It'll increase sprouts. especially for moose and some other things. It helps in revegetating the land as you can see by the amount of dead timber around here. That's desirable...[to]tryand get a mosaic of age classes beneficial to wildlife, grazing and watershed and for wildfire protection."The fire's beneficial aspects were touted ir:t several newspaper articles as well. The Anchorage Daily Press and theAnchorageTimesquoted Doshier and others explaining the ways in which fire is necessaryfor the ongoing health of wildlife. One article (phillips, 1991b)ta1ked about the expected blossomingofthe moose population as a resultofthis fire. The 1991 population in the Pothole Lake burn area,about two moose per square mile. is expected to increase significantly by the year 2000. The ICS team used the media to project a positive image. Wells.inhis interview. mentioned the benefits to moose. a sign that the media were successfulinreaching people and influencing their perceptions about this incident. Oneofthe most prevalent themes throughout the interviews was praise for how ICS was managing the fire. Sample comments included: Phil Todd: Painter: Perry: ''They didanexcellent job. Both professionally in getting the fire out, but also in keeping the community informed....."Ithink they were great.""Ithink the managementofthis fire ...has been real. real good.Ifyou get into the ICS system and become familiar with it, I think that you'll find that the way a fire is handled is basically the same...I think the managementofthis fire has gone real well. I think there's been some wise decisions made." Perry, a resource unit leader. alsotalkedabout the importanceofincluding local teams in the firefighting effort, since they know the land and the local people another example of ICS's sensitivity notonlytolocal needs, but alsotothe17


unique contribution that local residents and professionals can make. Klein, a crew bossofan Alaskan crew, talked about how well the ICS hierarchy works, in termsofthe emphasis on manageable spanofcontrol:"you know each squad boss andifthey're there, then their squad is there .. When we're in a position where we have to do something quickly or something like that I only havetocontrol two, three people. And they control threetofive's a great system, I think." Ground: "I think they didanexcellentjob...they did an excellent jobofbuilding the fIre breaks to keep it from gettinginthere." Chittendon: "I think this team's probably oneofthe best in the country." Doshier: ...the decisions weremade basedonthe best available information at the time. I think were good ones, were legitimate." Here, Doshier, the Kenai National Wildlife Refuge Manager, appearstobe defending the early decisiontolet the fire burn.D.Todd: "My experience with this's like it's's very pleasant and very professional." ICS clearly met Ms. Todd's expectationsofa disaster response organization. VanZee: ''The interagency cooperation was very good. The Incident Command System .. worked very, very well. That part went better than I expected." ICS exceeded the expectationsofVanZee. Liebersbach: "Well we use the Incident Command System in all our wildfIre's fit real well. We're organizedtodeal with fIres ...using that system and the people who were running the fire beforeuswere using itsothat made the transition very smooth coming in..." When not working onanincident with ICS, Liebersbach works for the BureauofLand ManagementinFairbanks. This praise for the ICS system and for the frrefighting efforts at Pothole Lake, comes from both residents and fxrefighting staff. The staff tendedtospeak highlyofthe system as a whole, while residents focused on the one experience they had with ICS -this fxre. The staffs respect for the system within which they work is likelytoenhance their loyaltytothat system, and consequently improve their job performance.EVALUATION OF THE POTHOLE LAKE FIREThe fire was formally evaluated internally by the Incident Command System and externally by the Alaska Fire Service, the agency which calledinthe Incident Command team. The intervieweesinformally evaluatedrcs'sperformance by commentingonitintheir interviews; these comments are discussed in the Qualitative Analysisofthe interviewtextThe Fire Overhead Performance Ratingisthe ICS internal evaluation. The performance ratings for Dave Liebersbach,inhis role as Incident Commander, are listed in Table 8.18


Therating numbers have the following meanings:o123Deficient. Does not meet minimum requirementsofthe individual element. Needs to Improve. Meets someormostofthe requirementsofthe individual element. Satisfactory. Employee meets all requirementsofthe individual element. Superior. Employee consistently exceeds the performance requirements.Theratingof"2"under Consideration for Personnel Welfare, is explained by attached comments suggesting that the Logistics Section "could have considered in this roadside fire situation cateringofmeals to crews and overhead," (Fire Overhead Performance Rating, 1991) and also suggesting the useofsuppression action showers for personal hygiene. All other comments are complimentary to the highest degree. Liebersbach and the team were especially commended for their attentiontosafety, their handlingofthe media, and their successful efforts at community relations: "It's imponant to note the excellent safety recordofthe incident. Considering the [hazardous environment], potentially very dangerous to personnel, the safety awareness level emphasis paid high dividends...Excellent attention was given to the media demandsofthe incident by the team. They build upon an established public information program already in place. Their understandingofthe political aspectsofa highly visible wildfire were well-substantiated..Community relations were excellent due to an organized effort by the team to leave the incident, local residents and facilities in better shape than they found them." These comments conflII11 the impression given by newspaper articles and interviews that attentiveness by the team to local needs, attentiveness by ICStothe needsofthe staff, and emphasis on communications all contributed to the successofthis incident.CONCLUSIONSA varietyoffactors contributed to the successofthe disaster response effort at the Pothole Lake Fire, including luck, useofthe Incident Command System, the natureofthe disaster, and the cooperationofthe townofCooper Landing and local agencies. In his interview, Liebersbach referred to this incident as an "easy" fire. He reponed that by the time his team had arrived, the fire was "not overly dynamic" and "was laying down." In 19


other words, the fIre had already taken aturnfor the better by the time the team arrived.Theweather also cooperated: temperatures. winds. and humidity provided favorable frrefighting conditions. Since there were no other incidents in the region atthat time. the team working on the Kenai Peninsula did not have to compete for emergency resources.Theinterviews. newspaper articles. and internal evaluation clearly identify theroleofthe IncidentCommandSystem in this success story.Theemphasis on safety was cited in the evaluation.Thegeneral sense that people hadofbeing taken care ofby ICS as well as their respect for the system helped to promote the positive emotionsofgroup cohesion. pride in their jobs, and pleasure in helping others.Itis likely that these positive emotions enhancejobperformance; this is a question that needs tobeaddressed in future research.Thepositive relationship that ICS had with the townofCooper Landing can be traced to ICS policies on communications. on unified command structure. and on minimizing disruption to the area in which an incident occurs. Carley and Harrald (1993) found that senior management presence is needed at a disaster site in order to shape media attitudes in a positive way. resulting in a more positive perceptionofdisaster responsebythe general population. Liebersbach and his section leaders appear to have played this key role. Media coverage and the interviews both demonstrate that ICS not only effectively communicated technical information. but also helped to ease people's fears and to comfort them. Another area for future research is to compare the effectsofcommunicating technical information with the effectsofcomforting people; what are the relative contributionsofthese two activities to a town's perceptionsofa disaster. relief team? Disaster plans sometimes are not followed "...because the planners were not the practitioners.Thepractitioners had no ownershipofthe plan.....(Carley and Harrald, 1993, p. 12-13). ICS policy addresses this issue in its requirementofa command structure, particularly in multi-jurisdictional incidents. Since key officials from involved agencies and jurisdictions participate in setting objectives for the incident, a feelingofcommitment to the successofthe plan flows back to the general public. Another ICS policy that appearedtoplayalarge role in the successofthis incident is the instructiontocause the least possible disruptiontoexisting systems. This instruction seemed to have a significant presence in Liebersbach's mind; he spoke extensively about it in his interview and interpreteditbroadly. He talked about being outofthe way when the tourist season started and about leaving the area cleaner than it had been when his team arrived. This ICS team seemed almost to consider themselves as guests in Cooper Landing. Their careful and considerate attitude were important factors in the successoftheir fxrefighting operation. Carley and Harrald (1993) have commented on the differences between hierarchical organizations that rely on standard operating procedure and teamsofpersonnel who are allowed to act primarily on the basisoftheir own experience.Theyfound that the advantagesofthe former typeoforganization include some emotional protection from the stressoftraumatic situations and greater resilience over the long term. These results appear to be borne out by the functioningofthe Incident Command System. Although someofthe membersofthis Class I team had worked together before, many had not. ICS's carefully20


defined detailed descriptionsofeach role needed in a fire incident allows a team ofpeople who have not had previous contacttoworletogether effectively and to communicate clearly. "Within hierarchies learning becomes embedded not only in personnel but in the relationships among personnel....(Carley and 1993, p. 19).ThePothole Lake rU'C was not catastrophic; no homes or lives were lost. The fU'C was experienced as beneficial to the environment, although this perception was dueinlarge parttoICS's expert handlingofpublic relations. Wildfires are not one-of-a-kind disasters; they occur regularly, and therefore emergency managers are willingtoplan extensively and allocate adequate resources for them. The teams that respondtothem have ample opportunity to test their organizational and fU'Cfighting skills in real-life emergencies, rather than only in training. Finally, credit for the successofthis incident goes tothe townofCooper Landing andtolocal agencies. Liebersbach commented in his interView that the town had already arranged meetings through which the ICS team could convey information, and he praised the workofthe planning group for the unified command structure. Town residents also volunteered for various support jobs within ICS. Because the sample in this case study is so small, we are more confident in the qualitative analysis results than in the quantitative analysis results, although the methodology itself is sound. Due to sample size, the main effect found for gender is not statistically significant.Wewould have expected role to be significant, but it wasn't. It is difficulttosay whether the results regarding role and gender are artifactsofthe sample size. These questions should be -investigated further. It iswonhnoting that the effort involved in performing the quantitative analysis is virtually the same regardlessofsample size. It would behoove future investigatorstosubstantially increase the sample size in order to get statistically significant results from the quantitative analysis. It is clear from the preponderanceofpositive adjusted E-scores that an overall tone of positive feeling surrounded this incident, whose success was independently corroborated by the Fire Overhead Performance Rating. This case study suggests that there is a link between the qualityofa disaster response effort and the emotional contentofinterviewees' speech.Themethodology (videotaping and transcribing interviews, coding the text, and analyzing the text quantitatively and qualitatively) is thorough and provides a varietyoftypesofresultstocompare.Werecommend using this methodology whenever possible in case studies researching emotion present during crises.REFERENCESCarley, K.M. (1990). "Content analysis."InAsher R.E. et al.,TheEncyclopediaofLanguageandLinguistics (Vol. 2, 725-730). Edinburgh, UK: Pergamon Press. Carley, K.M, Cohn, R.E., Harrald,JR..and Wallace, W.A. (1993). "Emotion and the21


American Red Cross's responsetoHurricane Hugo and the Lorna Prieta earthquake." Technical Report. Decision Sciences and Engineering Systems, Rensselaer Polytechnic Institute. Carley, K.M. and Harrald, J.R. (1993). "Organization Learning Under FlI"C: Theory and Practice." Quick Response Grant Report, Natural Hazards Research and Applications Information Center, UniversityofColorado. Carley, K.M., Harrald, J.R., and W.A. (1993) "Inter-organizational responsetocrises: The roleofemotions and information technologies." ProposaltoNational Science Foundation, Carnegie Mellon University. Cohn,R.E.and Wallace, W.A. (1991). "Field reportonPothole Lake nre inCooper Landing. Alaska." Decision Sciences and Engineering Systems, Rensselaer Polytechnic Institute. Ellis. B. (1991). "Firefighters hold ground; campsopen."TheAnchorage Times. May 30, 1991,AI.EndersJ.(1991). "Hot,dryspring sounds alarm for fire season."StateNews. June 3, 1991. Fire Narrative, Pothole Lake Fire, Fire Number 103108. Freedman, L. (1991). "Pothole Lake blaze threatens outdoorsman's legacy." Anchorage Daily News. May 30, 1991,Cl.HazardousMaterials Operations forFirstResponders (1990). Ashland, MA: International Society of FlI"C Service Instructors. Heise, D.R. (1977). "Social actionasthe control of affect" Behavioral Science. 22:163177.Heise, D.R. (1987). "Affect control theory: concepts and modeL"Journalof Mathematical Sociology.13(1-2):1-33.ICSstaff and Cooper Landing Residents, transcriptsofinterviews conducted by RuthE.Cohn. June 14. 1991. Decision Sciences and Engineering Systems, Rensselaer Polytechnic Institute. IncidentCommandSystemandStructuralFirefighting,The(1983). National Frre Academy. Kang. M.Wallace. W.A., and Carley,K.(1993). "Computer-aided content analysis." Working Paper. Decision Sciences and Engineering Systems, Rensselaer Polytechnic Institute. Kizzia, T. (1991). "Swift spread of flre forces new strategy." Anchorage Daily Press. May


29, 1991,AI.Liebersbach,D.(1993). Personal communication. Lien, L. (1993). Personal communication. NewJerseyHazmatEmergency Response Course (1990). New Jersey State Police OfficeofEmergency Management. Phillips,N.(1991). "Fire crews hope luckwilllinger:raincooperates, holding blaze back." Anchorage Daily News. May 29, 1991,BI.Phillips,N.(1991). "Kenai fIre may help wildlife." Anchorage Daily News. May 30,1991,BI.Phillips,N.(1991). "Peninsula fIre under control: campgrounds openatmidweek." Anchorage Daily News. June2,1991, AI. Reeves, S. (1991). "Pothole Lakefirerages, but rated only moderate." The Anchorage Times. June 2, 1991, B7. Spencer,G.(1991). "Local fIrefighters wishthey'd been called:' Anchorage DailyNews.June 3, 1991. Stone, PJ.,.Dunphy, .D.C., Smith, M.S., and Ogilvie,D.M.(1966).TheGeneral Inquirer: A ,Computer Approach to Content Analysis. Cambridge: MIT Press.


Table 1 Inteniewees N-.o.cripcJollCendw AMP ......A.,NNtIitttFnI.,.. M SIMINOTTeny"'*'-IrIcidMIe:--.,Lewel:I M SWl MGAISUPAltlefISIlIilIIFnI.... M SlIIIlNOTJilIl8ImeI M .,.MGRrSUPLyle0lineNIDftLogiIlaSec:tiona-.. M S1aIIMGAlSUPOMe-n.rICeMiReluge.....,. M $WIMGAIS\JP00rlIlhyToddPIwwin;IUPPOIt.,.....,. F AeaidMNOT P. GnMlniCooperI.M*lo--. F AeaidMNOT 0 ......1quiIIFnf"'. F S1aIINOTo..JanclI0NiIi0n1UPel-..... M .,.MGRrSUPJohnSmilhInl:idMIeomm.nd&q:Dft M SlaIlNOT C ....bou M Slat! MGPlSUP Ron KnowInSeCIiona-.. M SlIIIl MGPlSUPJakeL.,..CoopI.M*lo....... M NOT0... Lieberat.dlInl:idMICommMderLewel1 M SlIIIl MGPlSUPL.indMyLienUniI M S1aIIMGRrSUPNr.r::yLuebbertOemobiizaliDnUniIL.__ F SlaIlMGRrSUPNeilMarDnbInkaMel.-logill M SlaIlNOTManyMelMitty (two) M SIIIIf NOT MonaPUll.,Cooper!Aldin;--F Rea.den NOT BUlChPenyR_UtilL..der M SIaIIMGRlSUP Phil Todd..R.aidenl NOT MarvinRobenaonPlanningSecIiona-....SIIIIfMGRrSUPCtySlal StordalFnfigllw F SlaIl NOT Stnokejumperldemollilizaliml..SIaIt NOT BrweVanZeeNcionalFe,....SIaItMGR'SUPGeorpeW.I.CommandPDSIRegioa_ M SlaIl NOT


Table 2 Word Usage StatisticsTotal number of base and discrete910words used Number of base words used313Numberofwords with exact275matches to Heise's list Numberofwords where ratingsfor25similar Heise words were used Number of words with no Heise13word similar enough to useTable 3Most Frequently Used Words Ten most frequently Heise E-rating: used words and their counts:MF1out404--2like389 2.24 2.48 3work308 0.07 0.994 good220 2.48 2.66 5want141--6kind138 (removed)(removed) 7need125 1.38 1.70 8mean76I (removed)(removed) 9train70I1.24 1.14 10help67I2.17 2.34


Table 4 Emotion ScoresName Raw Score Word Count Adj. E-score Roger Arnatln66.07 2520 2.62Terry Anderson63.2 1276 4.95Albert Smith17.95 846 2.12Jim Barnett133.64 2968 4.50Lyle Chittendon186.82 5572 3.35Dan Doshier50.56 2328 2.17Dorothy Todd303.1584085.13Pat Ground5.08 219 3.76Darla Hasselquist48.34 1206 5.57Dave Jandt100.44154 3.61John Smith11.n399 2.32Dennis Klein168.26 51724.01Ron Knowles20.26 1239 2.42Jake Lenarik-9.984912.95Dave Liebersbach185.48 2736 3.25Undsey Lien43.8 1987 1.64Nancy Luebbert34.41 27004.60Neil Maronbanks15.95 656-2.03Marty and Marty (two)97.57 3476 6.78Mona Painter46.5225083.17Butch Perry50.46118512.20Phil Todd47.72 930 1.27Marvin Robertson37.5910012.16Crystal Stordal392.4870516.35Mike Tupper95.53 20781 2.43Bruce VanZee93.429511.85George Wells118.98 1874 4.26


Table5 H)'potheses andResults VARIABLE D.F.FVAlUEPVAlUEGenderHo : Pr = P. I7.47 .0128H,: Pr:f:Pm RoleHo : )1.=P, I1.09 .3091H,: )1.:f:p, PositionH o : J.l-=P_ I0.03 .8707H,: Pm.:f:PnmaPF Mean adjusted E-score for females Pm= Mean adjusted E-score for males PI= Mean adjusted E-score for staff p,= Mean adjusted E-score for residents Pma= Mean adjusted E-score for mgrllsup staff P_= Mean adjusted E-score for non-mgrllsup staffTable6 Univariate Statistics byCellGender Role PositionNMean StdOev Min Max Not Managerial22.111.191.27 2.95Resident Managerial0----Not Managerial6 1.95 2.10 -2.03 4.26Male Staff Managerial123.01 1.091.64 4.95Not Managerial34.021.013.17 5.13Resident Managerial0----Not Managerial2 5.96 .551 5.57 6.35Female Staff Managerial1 4.60--4.604.60


Table 7 Emotion-Related ThemesTheme#Pride/pleasure/excitement about theirworkor about this fire12Praise for ICS's firefighting efforts11Good communications, the importance of communications7Sense of family, group cohesion, group dynamics5Feeling of being taken care01bythe ICS organization4Importance01good training, firefightingasa learning experience4Altruism, sense of protectiveness toward the region or town3The Pothole Lake fireasbeneficial3Sense of competition, elitism2Table 8 Fire Overhead Performance RatingRatingFactors:Rating: Knowledge of the job3Attention to Fire Fundamentals3Ability to Obtain Performance3Practice Race and Sex Equality3Attitude3Initiative3Consideration for Personnel Welfare2Obtain Necessary Equipment and Supplies3Physical Ability for the Job3Safety3Summary3