micronesian - charles sturt universitymicronesia.csu.edu.au/mjhss/issue2004/mjhss2004_06.pdf ·...

This is a peer reviewed contribution. Received: 19 Nov 2004; Revised 15 Dec 2005; Accepted: 27 Dec 2004.© Micronesian Journal of the Humanities and Social Sciences ISSN 1449-7336Letao Publishing, PO Box 3080, Albury NSW, Australia 64

MICRONESIANJOURNAL OF THE HUMANITIES AND SOCIAL SCIENCES

Vol. 3, nº 1-2 December 2004

DESIGN AND APPLICATION OF ON-LINEQUESTIONNAIRES: EXPERIENCES FROM MICRONESIA

Jon G. O’Neill & Dirk H. R. SpennemannCharles Sturt University

The administration of mail-out questionnaires in Micronesia is plagued by the tyrannies of spatial and cultural distances:Micronesians live dispersed throughout the islands of Micronesia and various locales in the island and mainland USA.They have also been described as people preferring oral exchanges over written communication. While the administra-tion of paper-based questionnaires is challenging, observations of a digital divide in the Pacific make the use of on-linequestionnaires a daunting proposition. This paper describes the development of the survey instrument and how it wasreceived by the population surveyed. Advantages and disadvantages of the methodology are discussed.

INTRODUCTIONMicronesia is located in the northern PacificOcean and forms part of Wendt’s imaginedOceania. It is a vast area of ocean that stretchesover approximately 27 degrees of latitude and44 degrees of longitude making it about 5,000kilometers long and 3,000 wide. Despite con-sisting of almost 1.5 million square kilometersof ocean, there are only about 2,600 squarekilometers of land distributed between ap-proximately 2,200 islands. Few of them aremore than a hectare or two in size, few aremore than two meters above sea level, manyhave little or no fresh water, and most are un-inhabited. Because of their physical isolation,the Micronesian islands were not settled untillater in human history. For the same reasons,their first inhabitants were among the last to beaffected by European colonialism, and now thenation-states of Micronesia are among the lastto gain national independence.

Throughout Micronesia, responsibility formanaging the dynamic cultures being con-structed by these peoples has traditionally beenin the hands of the chiefs. It was one of theirparticular obligations and responsibilities andcould not be ignored (cf. Carucci 1997). Overseveral hundreds of years in some parts ofMicronesia, and over a shorter period in others,colonial powers gradually eroded the authorityof traditional forms of government.

During the late twentieth century, theUnited States National Park Service imple-mented deliberate policies calculated to slowprocesses of change within the several culturesof Micronesia and preserve aspects of thosecultures. This has continued in conjunctionwith the Historic Preservation Offices of eachof the political entities that have emerged fromthe Trust Territories of the Pacific Islands, i.e.the Republic of the Marshall Islands, the Fed-erated States of Micronesia, the Republic ofPalau and the Commonwealth of the Northern

Design and Application of On-Line Questionnaires: Experiences from Micronesia 65

Marianas; Guam, a former US possession andnow a territory, also needs to be mentionedhere. In practical terms this has contributed tothe development of a preservation ethos that isvery similar to that practiced within the UnitedStates and other western nations—a conserva-tion ethos that tends to emphasize tangibleheritage at the expense of intangible heritage.Significant resources in terms of time, energyand ‘hard’ money were being used to preserveheritage material in projects that funding sour-ces (almost always external and predominantlyfrom the USA) considered appropriate, such asWorld War II battlefields and relics, buildingsand other constructions (cf. Adams et al 1997;Christiansen 1994a–d; Denfeldt 1979a–b; 1981;Hezel & Graham, 1997). Exceptions, outsidethe realm of straight archaeology, are few (cf.Parker, 1994; Galvan 1998).

The tension between the preservation ofWWII period heritage items, which after all arerestricted to a very short period of history, andthe remainder of the tangible heritage has beenexplored by Spennemann (1992a-b). Duringearlier research it was shown that not allMicronesian aspirations or preferences forpreservation of their unique cultures and heri-tage were being satisfied (Spennemann 1992a;O’Neill 2000; O’Neill & Spennemann 2002).Further research was needed to determine In-digenous Micronesian perceptions of heritagepreservation—what elements are especiallyvalued, which should be preserved and howthey should be preserved, what prioritiesshould be established.

Methods used to collect these data were in-depth interviews with key stakeholders and de-cision makers, as well as a survey questionnairethat polled both adults and school children. Toaccommodate the geographical and cultural re-alities in Micronesia, the questionnaire em-ployed a variety of delivery methodologies.One of these was a computer-mediated surveyinstrument, an on-line questionnaire

This paper describes the cultural back-ground and conditions which influenced thedevelopment of the instrument and discussesthe experiences with this questionnaire, in par-ticular the successes and short-comings of de-livering such a questionnaire on-line.

CULTURAL CHALLENGESMicronesia is not only a collection of islandsand atolls that are widely distributed geo-graphically, it is also comprised of many variedand unique cultures. Within each of them, wehave to consider the dynamics betweennational centers and the out-lying districts orislands; dichotomies of power, rank and status;and differences between traditional values andmodern commercial realities and influences.Historically, these communities had a culturethat was largely non-confrontational and wheredecision-making processes were strongly basedon discussion and achieving consensus. Thistradition of consensus and non-confrontationis still strongly entrenched in all of the culturesof Micronesia. Personal opinion is expressedwhen sought, but rarely volunteered.

Micronesian cultures were also oral cultures.Even today, after many decades of accultura-tion and colonially provided schooling, theemphasis upon written records is far less a partof their cultures than in the ‘western’ world.Thus opinion is provided when prompted ininter-personal communication, such as aninterview, but withheld or hesitantly providedvia written means.

As these features apply generally across allMicronesian cultures, and much of the Pacificin general, written mail-out questionnaires,which are the standard mode of informationgathering in the western world, cannot be con-sidered to be an ideal mode for soliciting in-formation in Micronesia. Indeed, wherequestionnaires have been used in the Pacific,response rates of mail-out questionnaires haveoften been very low. A few examples may suf-fice.

In a mail out targeting Pacific Islander stu-dents at a ‘large multi-ethnic, urban polytechniclocated in an area of low socio-economic sta-tus’ in New Zealand, Rolleston and Anderson(2004) obtained a response rate of only 9%,despite mail-out follow ups. A nutrition surveytargeting Pacific Islanders residing in NewZealand showed a response rate of less than20% (Wham & Worsley 2003). Other studiesdrawn from a variety of research fields alsounderscore the lower response rate among Pa-cific Islanders (Saphira & Glover 2000).


It can be posited, that the response rateshould be higher among those sectors that ac-tually manage and use information technology.The reality, however, does not confirm this: Aquestionnaire sent in 1995 to 137 managersand users of the, then, PAKTOK communica-tions network had a response rate of 24%(Ward & Spennemann 2001). A questionnaireto estimate the number of Pacific Islands peo-ple having access to the WWW, sent to 85 in-dividuals in 1999, had a response rate of 48%(Zwimpfer Communications 1999), which isacceptable by any standard. The 2002 repeatsurvey sent to 161 individuals, however, had aresponse rate of only 23%, despite 83% of thepotential respondents being called personallyby phone in a follow-up to the mailing to en-courage completion of the questionnaire(Zwimpfer Communications 2002).

Despite these problems, a questionnaire-based approach was seen as the preferredmeans of data collection, with the inherentproblems to be overcome by select means ofadministering the survey instrument.

QUESTIONNAIRE DEVELOPMENTAs Gillham (2000) suggests, constructing aquestionnaire is easy, but developing one thatwill produce useful data is a much more diffi-cult task. The process of producing a well-designed and useful questionnaire is “a pro-longed and arduous intellectual exercise” andits purpose is “aimed at precision, logic-tightness and efficiency” (Oppenheim (1982,pp. 3-4). Questionnaires have an inherent po-tential for bi-directional misunderstanding.This is particularly true when complex subjectsor issues are involved. As many researchershave noted (see Frazer and Lawley, 2000; Gill-ham 2000; Brady et al 2003) comprehension byparticipants for whom the questionnaire lan-guage is not their primary or preferred lan-guage may present difficulties. Similardifficulties may result from cultural differenceswhen questionnaires originate in one culturebut are completed by respondents from an-other culture. This issue may be exacerbatedwhen a variety of different cultures, ethnicitiesand languages are involved.

Choices and compromises in survey designare unavoidable. For example, what contactand response modes will be utilized? Howmany questions should there be and howshould they be worded? What methodology orcombination of methodologies should be usedand what levels of technology are appropriate?Should Questionnaires be paper-based anddistributed (by mail or by hand), completedand collected manually or can computer-aidedfeatures be added? How may people best beencouraged to participate? Should incentives beoffered and if so, what kind? Answers to ques-tions such as these must also be evaluated inthe light of resources available to the re-searcher and the nations chosen (cost, time,available technology, geographic distribution,etc).

The option of distributing and collectingquestionnaires by mail was discarded for sev-eral reasons, including:• Inequity of opportunity—the geography

of the region with islands and atolls (par-ticularly ‘outer’ districts) results in popu-lation centers being widely dispersed withfrequently irregular and occasionally unre-liable transportation between them.

• Expense—Mailing substantial numbers ofquestionnaires (in excess of 500) to indi-vidual participants, having them com-pleted and individually returned by airmail was prohibitive.

• Anticipated Response Times—Postalservices within Micronesia and betweenMicronesia and the outside world are no-toriously slow and unreliable: parcels ofcompleted questionnaires took fourmonths (Rota to Australia) and fivemonths (Kwajalein to Australia).

• Promotion—It would have been moredifficult to effectively promote the Ques-tionnaire without personal interaction al-lowing participants to ask questions aboutthe research and its publication.

Completing forms can be problematic fortraditionally oral cultures which are not neces-sarily familiar with the processional, top tobottom, left to right model of information flowthat is common in western cultures and whichis based on a different cognitive method.


Internet based questionnaire deliveryThe use of the internet as a survey tool hasbeen around since the mid to late 1990s (cf.Carver et al 1998 and references therein). Ad-vantages for internet-based questionnaires in-clude aspects such as social anonymity of therespondents (cf. Joinson 1999). Bowker andDillman (2000) comment that traditional ques-tionnaires “implicitly assume a cultural modeof cognition in that they rely heavily on the ex-pectation that readers will employ a conven-tional ‘paper-logic’ while filling out the websurvey”. This difficulty is exacerbated whenquestionnaires are lengthy, involve complexissues, and are written in a foreign language.Other studies, however, have shown that re-sponses between on-line and paper basedquestionnaire are no different, especially whenboth are drawn from the same population (inthis case self-nominated via the internet) (Ritteret al 2004).

Areas of questionsInformation sought in the questionnaires con-centrated particularly on the following threeissues:1) How different strata within Micronesian

societies (age, education, ‘central’ and‘outer’ population, ethnicity etc) view heri-tage preservation and the importance andrelevance of their traditional cultures totheir societies of today.

2) What elements of their heritage Micro-nesians themselves consider to be mostworthy of protection and management.

3) Whether the effects of education, em-powerment and improved communicationon perceptions of heritage values can bemeasured.It was deemed desirable to assess these

questions by surveying across the strata of ageand education. To this end, two base question-naires were prepared, one targeted to schoolchildren in their later years in Pri-mary/Elementary School and the other toHigh School, College or University students aswell as other adults. Both collected data byvarious strata and the questions comprisedmultiple-choice and narrative components.

Questions were presented in clear and con-cise terms using simple English and non-complex expressions. Although multiple-choicequestions formed the bulk of the questions (forspeed and commonality) opportunities weregiven at strategic points for respondents to usetheir own words to express their feelings. Thefirst such opportunity was on the first page ofquestions so that interest was maintained orgenerated. From first invitation to the finalthank-you the importance of each individuals’own opinions, values, perceptions and partici-pation were strongly emphasized.

To evaluate comprehension, suitability andtime taken to complete them, the question-naires were pre-tested in Australia with adultand child respondents from non-Indigenousand Indigenous (Aboriginal and Torres StraitIslander) communities. Apart from requiringminor wording changes to several questions,the results of the pre-test indicated that thequestionnaires should be easily comprehendedand that the average time taken to completethem (approximately 15 minutes) was satisfac-tory.

Sample PopulationsIt was decided to employ direct administrationof the questionnaire in ‘captive audience’ situa-tions (class rooms, meetings etc) and a webbased questionnaire. Used together, both stylesof questionnaire were intended to ensure thatmost strata of Micronesian society had oppor-tunity to participate in the survey. A significantproportion of the populations of Micronesiannation/states live ‘off-island’ in the UnitedStates and elsewhere (Spennemann 2003). Fre-quently, expatriate Micronesians are significantsources of funds to other family members stillin Micronesia, and it can be posited that theiropinions will also have considerable influence.Anecdotal evidence suggests Micronesians liv-ing overseas are generally more highly edu-cated, have a more ‘western-oriented’ workethic, and are less concerned with the inherentconflict between economic development andcultural preservation. In some ways this diasporamay be regarded as part of a Micronesian edu-cated “elite” and this is sometimes consideredto be the case.


Figure 1. Screenshot of Yokwe.net showing the announcement of the questionnaire

However, such a presumption may be un-justified (at least in terms of heritage preserva-tion), and the questionnaire may providestronger evidence than exists currently. Cer-tainly, without the Internet-based questionnairethose Micronesians living overseas would nothave had opportunity to contribute to the re-search.

PromotionFor the paper-based questionnaire, severalmethods of promotion were adopted includingpersonal contact, conference meetings, meet-ings with Historic Preservation agencies in theUnited States and Micronesia, e-mail, tele-phone conversations, public lectures, guestlectures and presentations at schools, collegesand the University of Guam.

The internet-based questionnaire was ad-vertised by personal contact at conferences,public presentations and lectures, invitations to

participate were distributed by email and ‘chat-room’ messages, as well as notices being postedon a relevant community Internet portal (figure1). It was promoted within Micronesian com-munities with access to the Internet, and com-pleted and submitted electronically. IndigenousMicronesians were the primary target, whetherliving in Micronesia or elsewhere (such as theUnited States, Australia or New Zealand), butprovision to participate was also made for non-Micronesians who were interested in Micro-nesian heritage.

QUESTIONS OF DESIGNSeveral issues were recognized as requiringparticular care in designing a format for thequestionnaires. Both paper-based and web-based questionnaires had to include the samequestions so that they were equivalent and datafrom both could be easily combined and ana-lyzed separately and collectively.


It is important that surveys be structured sothat the primary questions are asked early. Thisencourages respondents to continue becausethey begin by answering the core material ra-ther than become bored with questions thatmay not appear to be truly pertinent to themain research (Websurveyor Corporation2000). Early questions also tend to set the tonefor a survey. If not carefully worded, appropri-ately grouped and well positioned in the ques-tionnaire they can give an unfavorableimpression of the research which will probablyresult in lower response rates. Questionsranged from the regional Micronesian level,through national, state, and home dis-trict/island levels to the personal level thuscontributing to a sense of personal involve-ment and ownership of the research.

With that in mind, both questionnaires, pa-per-based and web-based, were structured insuch a way that core questions (those that weremore important in terms of the main purposeof the research) were grouped at the beginningand questions that were less important werepositioned towards the end. Although the firstpage contained some demographic questions,they were structured to encourage respondentsto develop a personal involvement with thequestionnaire by association. Thus the ques-tionnaire began with questions that allowedeach respondent to associate with the researchthrough national, ethnic, cultural and ‘home-district’ linkages. Other demographic questions(usually considered boring or even intrusive)were left to the penultimate group with thefinal group containing questions relating to theresearch and the questionnaire itself.

Structural and Technical IssuesWhile the paper-based and web-based ques-tionnaires contained the same questions,structural adjustments needed to be made forthe web-based questionnaire to function. Theremainder of the paper will concentrate

Computer-mediated questionnaires havecontributed some issues of concern related totheir use of a new medium. Dillman et al.(1998) show there are important questions re-garding the use of “fancy” or “plain” question-naires that must be considered before a final

design is established. Questionnaires designedto be ‘fancy’ generally require greater band-width or they are slow to load and thus cause adelay, which may induce the user to abandonthe process. They often also require a higherfunctionality of available IT equipment, andmore current software than those designed tobe “plain”.

One of the most critical early tasks was toclearly define the target audience and the tech-nical capabilities of the service available tothem. In the particular circumstances ofMicronesia, this included the level of technol-ogy to which participants could be reasonablyexpected to have access and the probable levelsof their ‘web-literacy’. A report produced byUN-ESCAP (2001) stated, for example, that atthat time the IT infrastructure in place in theFederated States of Micronesia was only “aver-age”. While 8,000 to 10,000 computers wereinstalled throughout the Federation, only 1,800were connected to the Internet. Elsewhere,Spennemann (2004a) has outlined the DigitalDivides prevalent in the Pacific and has shownthat most of Micronesia, with the exception ofGuam and the CNMI, suffers from slow con-nections, high costs and low penetration of ITinfrastructure.

Thus, most frequently, the lowest commondenominator must become the limiting factorin using such technologies. The Internet-basedquestionnaire was therefore designed conser-vatively to permit the widest possible audienceparticipation. Minimal graphics were used as itwas important that the speed of loading foreach page was as high as possible. StandardHTML (version four) was used exclusively tobuild the pages and more advanced HTMLfeatures such as “style sheets”, “linked win-dows” and “frames” were specifically excludedfrom the design. To ensure maximum com-patibility with the widest possible variety ofsoftware and hardware, no proprietary Net-scape® or Microsoft® extensions or featureswere used. The questionnaire could thereforebe aptly described as “plain” (Dillman et al1998).

While respondents to the paper-basedquestionnaire had the entire questionnaire infront of them, and thus could make a decision


to participate or not, web-based questionnairesare not as clear, which might induce partici-pants to abandon the process half-waythrough. Couper et al (2002) demonstrated thatdisplaying multiple questions on a screen atonce rather than singly or as one long ques-tionnaire, reduced completion time, producedfewer non-answered questions and raised levelsof consistency in answers. Thus, rather thanpresent the questionnaire as a single monolithicwebpage, it was separated into logical sectionsand presented on-screen one section at a time,with each section containing multiple questions(figure 2). When each page was completed, theanswers from that page were posted separatelyto a receiving directory. A non-identifying codewas allocated to each individual access andlinked with each questionnaire page completed.This permitted all sections and answers com-pleted to be collected progressively rather thanonly at the completion of the final page.

A total of fifty-eight separate web pageswere prepared and linked to permit a variety ofpathways depending on answers to key ques-tions. In this way complications arising frominappropriate, repetitive or unnecessary ques-tions were avoided. The first page displayedconsisted of a brief description of the research,an invitation to participate and request forpermission to use the information provided. Italso included links to other pages containingmore detailed descriptions of the research andethical rules under which this research wouldbe conducted. Respondents were offered theopportunity to accept or decline the invitationto participate in the research. If the invitationwas not accepted, a further page was displayedwhich discussed the importance of participa-tion further and asked respondents to recon-sider their decision. If this was accepted thenrespondents were returned to the Question-naire and the first question page was displayed.When the questionnaire was completed and thefinal answers submitted, a special “Thank-you”page was displayed which included instructionsfor accessing the analyzed data and reports asthey became available.

There were seven main pathways:• Adult Micronesians living in Micronesia

• Adult Micronesians not living in Micro-nesia

• Adult non-Micronesians• Elementary School Students attending

school in Micronesia• Elementary School Students attending

school outside of Micronesia• High School Students attending school in

Micronesia.• High School Students attending school

outside of MicronesiaFrom the point of acceptance, respondents

were directed to different questionnairestreams depending on their replies to specifickey questions (Figure 2). Children attendingElementary school would see four pages con-sisting of 53 questions, High school studentswould see seven pages (124 questions), non-Micronesian adults would see seven pages (110questions), and Micronesian adults would seenine pages (167 questions).

Ease of ReadingThe education systems in each of the politicalentities that have emerged from the formerTrust Territory of the Pacific Islands areheavily based on the United States educationsystem. US school curricula are often used tosuch an extent that American subjects (such ashistory and geography) are frequently taught atthe expense of more culturally relevant localhistory (pers.obs; Micronesia is not unique inthat regard (Spennemann & Meyenn 1997).English is a core subject from an early stage inall Micronesian education programs, but itmust be acknowledged that in many instancesit is neither the ‘first’ nor the preferred lan-guage. Consequently, questions had to bestructured very carefully so they were ex-pressed in as simple a manner as possible andcompound questions and jargon were rigor-ously avoided.

Readability tests are simple techniques forpredicting the reading grade level required ofaverage readers that they might understandwritten material. They can measure the struc-tural difficulty (sentence structure, idea densityand vocabulary) of the text of a questionnairebut cannot guarantee its effectiveness (Flesch1951). The standard Flesch-Kincaid Grade


Level test (which provides an estimate of theU.S. Grade level required for reading a givendocument) and Flesch Reading Ease tests(both are automated and available in MSWord™) were used to estimate readability of

the Questionnaires. The paper-based question-naires were tested in their entirety, while eachpage of the computer-mediated Internet-basedquestionnaires was tested separately giving arange for each questionnaire (Table 1).

124Questions

124Questions

53Questions

53Questions

167Questions

167Questions

110Questions

inMicronesia

not inMicronesia

Micronesiannot in

Micronesia

NonMicronesian

Micronesianin

Micronesia

inMicronesia

not inMicronesia

DECISIONPOINT

ElementarySchoolStudent

DECISIONPOINT

AdultHigh

SchoolStudent

DECISIONPOINT

INVITATIONResearchDetails

PLEASERECONSIDER

END

Permissionto use data

NO

YES

YES

DECISIONPOINT

Figure 2. Flowchart of the on-line Questionnaire


Both sets of results indicate that the ‘reada-bility’ of both paper-based and Internet ques-tionnaires are appropriate for their targetedgroups. It is considered that a grade-level resultof 9 is generally an acceptable level of difficultyfor questions posed in a survey to adults. In theInternet questionnaire, two pages of the Adultpathway included a high percentage of state-ments not written as complete sentences but asa series of responses to a common sentencebeginning. Consequently, when standard com-puter software tests are used to assess theirreadability they appear to be difficult (Flesch10.9 and 11.2). To more properly evaluate theirreadability, the statements were temporarily re-assembled to full sentences and reassessed.When this was done, both pages scored signifi-cantly higher on the Flesch Reading Ease scaleand lower on the Flesch-Kincaid Grade Level.

Table 1 Ease of Reading

Flesch-KinkaidGradeLevel

FleschReading

Ease

Paper-Based QuestionnairesElementary School students 2.2 90.7Adults/High School students 8.0 55.8

Computer-Mediated QuestionnairesElementary School students 2.3–3.9 79.7–93.9High School students 4.5–8.8 54.1–79.2

Adults(11.2)

4.9–8.8(34.8)

53.7–76.0

CALCULATING THE RESPONSE RATEAlthough the total number of responses to thisInternet questionnaire is not high, the responserate has exceeded expectations. In paper-basedquestionnaires the response rate can be easilymeasured in terms of questionnaires returnedin relation to the total number distributed. Thenumber of responses to an online deliveredquestionnaire, however, is potentially enor-mous and, in practical terms, is not readilymeasurable. While actual responses can beeasily totaled, a potential uncertainty remains asto how many people accessed the questionnairebut chose not to respond. To overcome thisproblem, the design of the online questionnairepurposefully included the already mentioned

introductory page. An analysis of the server logfiles permits us to quantify the number of peo-ple actually accessing this file and thus thequestionnaire (to maintain confidentiality, theserver log file was analysed by DHRS while thequestionnaire response content was analyzedby J O’N; confidentiality is not a concern inthis paper as it reports only on process).

The questionnaire went ‘live’ on 29November 2002 and remained accessible until24 February 2004. During that period 227 ‘hits’were recorded on the first or invitation page, apage that described the research, discussed theimportance of participation and sought permis-sion to use the information provided. An IPaddress analysis of these, however, showed that44 of these were created by web crawler robotsoperated by various internet search engines(Google, AltaVista and AskJeeves) which hadpicked up the page reference from the siteslinking to the questionnaire. Thus only 183‘hits’ were genuine requests for the page. Over50% of these occurred during the first fourweeks (Figure 3).

0

10

20

30

40

50

60

70

80

90

100

Nov–02

Dec–02

Jan–03

Feb–03

Mar–03

Apr–03

May–03

Jun–03

Jul–03

Aug–03

Sep–03

Oct–03

Nov–03

Dec–03

Jan–04

Feb–04

Figure 3. Interest in the Questionnaire November2002 to February 2004

As the invitation page, as well as subsequentquestionnaire pages, required a ‘post’ action bya user (ie. clicking on a button), the search en-gine robots could not proceed further. Exactlyone hundred human participants continued tothe next page. The initial response rate wastherefore 54.6% percent. Table 2 sets out theflow of respondents through the first sevensteps/pages of the questionnaire. As the ques-tionnaires were of differing length, we cannot


compare the retention rate beyond this step. Itbecomes evident that the percentage of re-spondents decreases as the questionnaire pro-gressed. Fifty-four respondents completed thequestionnaire in one of its configurations. Thisis a 54% retention rate of those who startedand an overall response rate of 29.5%.

Table 2. Flow of respondents through the questionnaire

n PercentView Page 183 100.0Start Questionnaire 100 54.6Step 2 97 53.0Step 3 92 50.3Step 4 89 48.6Step 5 72 39.3Step 6 69 37.7Step 7 63 34.4Complete 54 29.5

The computer-mediated questionnaire dif-fers fundamentally from paper based question-naires in one aspect: not all respondents ofpaper-based questionnaires who terminate theprocess before reaching the end actually handin their incomplete forms, and thus reduce theresponse rate. The computer-mediated ques-tionnaire, with its incremental completion ofthe questionnaire components, however, re-cords all responses until that page where re-spondents actually abandon the process. Thisresults in a differential response rate favoringquestions asked early in the questionnaire. Ifwe consider an overall response rate of about40% as acceptable, for example, then all ques-tions asked up to step 5 are deemed above thecut off, while subsequent pages are below.

PERCEIVED AND REAL TIME TAKEN TOCOMPLETE THE QUESTIONNAIREOne of the major concerns in questionnairedesign is the time it takes to complete a ques-tionnaire. On-line questionnaires allow to as-sess the time it takes, as each posting of acompleted segment of the questionnaire is re-corded in the server logs. It needs to be notedthat elapsed time does not necessarily equate tothe time spent answering the questionnaire butmay include time doing other tasks away fromthe computer. That is, a questionnaire may bestarted, but abandoned somewhere in the mid-dle, to be picked up and completed at a later

point, while still connected. The same is pos-sible for paper-based questionnaires, but nor-mally undetectable.

The true elapsed time taken to completequestionnaires could be calculated for 98 re-spondents (Table 3). Those who completed theentire questionnaire did so in an average timeof 25 minutes and 24 seconds, with the maxi-mum elapsed time being 1 hour 48 minutesand 48 seconds. Not all, however, completedthe questionnaire but terminated early. The av-erage time elapsed until completion of the lastpage before termination allows us to assess the‘patience’ of those not willing to complete: justunder seven minutes. Yet, the maximum of 1hour 23 minutes and 15 seconds also showsthat this has great variation.

Table 3 Elapsed time (h:m:s) to complete/terminatequestionnaires

Average Stddev Max nCompleted 25:24 19:09 1:48:48 54Terminated 6:58 13:54 1:23:15 44

Despite the fact that elapsed time in fourinstances was greater than one hour and thelongest elapsed time was more than one andthree-quarter hours, only seven participantsfailed to complete the questionnaire after beingconnected for 15 minutes. Seventy-seven per-cent of those who did complete it did so inthirty minutes or less.

0:00

0:15

0:30

0:45

1:00

1:15

1:30

1:45

2:00

1 2 3 4 5 6 7 8 9 10 11 12 13

– 1s AVG + 1s Maximum

Figure 4. Average cumulative time taken (h:mm) forthe completion of consecutive components of the on-line

questionnaire.

As expected the average cumulative timetaken to complete various steps of the ques-


tionnaires is a linear process (Figure 4). An ex-ception is the curve plotting the maximum timespent, which shows two discrete steps. Sug-gesting that in that case the questionnaire wastemporarily abandoned.

Table 4 Comparison of elapsed time for completed orterminated questionnaires

Time (min) Completed (%) Terminated (%) n0-5 1.85 70.45 326-10 11.11 11.36 1111-15 22.22 4.55 1416-30 42.59 9.09 2730-45 9.26 2.27 646-60 7.41 0.00 460+ 5.56 2.27 4TOTAL 54 44 98

It is worth observing that of those whofailed to complete the questionnaire aftercommencing it, seventy percent did so withinonly five minutes of starting (As expected theaverage cumulative time taken to completevarious steps of the questionnaires is a linearprocess (Figure 4). An exception is the curveplotting the maximum time spent, which showstwo discrete steps. Suggesting that in that casethe questionnaire was temporarily abandoned.

Table 4). That only 13 (13.5%) chose toterminate after five minutes and the fact that noexternally derived inducement (such as mon-etary) was offered to encourage participationstrongly suggests that those who completed thesecond page felt sufficiently motivated to con-tinue regardless of the size and complexity ofthe questionnaire. Given an opportunity tocomment or make suggestions about the re-search and questionnaire, participants madecomments like: “Good survey. I look forwardto receiving a copy of the results”, “I will lookforward to results of your research”, “I lookforward to seeing the results on your webpage”, and “I believe this is good thing” (sic).Such comments suggest that participants feltpersonally involved and developed a sense of‘connection’ to the research.

Forty-one participants responded to thequestion “How much time do you think it tookto complete this questionnaire?” We can cor-relate their perception with the actual elapsedtime as indicated by the computer logs (Table

5). Overall, 56.1% underestimated the actualtime it had taken them, while 26.8% were ap-proximately correct. Only 17.1% overestimatedtheir time investment. This indicates that therespondents were, on average, so engrossed inthe questionnaire that they forgot about time.

Table 5 Correlation between perceived and actual time(in minutes) to complete questionnaires

Actual Time

perceived time

Lessthan10

10to20

20to30

Morethan30 Total

Less than 10 2 5 3 1 1110 to 20 2 7 6 6 2120 to 30 0 4 0 2 6More than 30 0 1 0 2 3TOTAL 4 17 9 11 41

ORIGIN OF RESPONDENTSOne of the major advantages of on-line ques-tionnaires is the relative social anonymity ofthe respondent (cf Joinson 1999). Thus, whilethe audience of paper –based questionnairescan be assessed correctly by means of distribu-tion, this is not the case for internet-basedquestionnaires. Responses can, literally, comefrom the world over. This can be assessed intwo ways: via responses to questions in thesurvey instrument itself and by an analysis ofthe computer log files.

The survey instrument asked participants tostate their ethnicity as well as their (current)place of residence. The response by ethnicityshows 44.7% of responses came from Micro-nesians, 32.8% from Americans, 13.8% fromAustralians and New Zealanders (Table 6).

Table 7 sets out the country of origin ofthose who called up the initial web page. Thiscaptures both actual respondents and inter-ested individuals. Given privacy considerations,the IP addresses were only correlated to coun-try and nature of the IP address by broad cate-gory. The US domains dominate the sample byfar (78.1%), while truly Micronesian domainsare few (3.84). However, the actual physical lo-cation of .com domains is no longer predict-able as such domains are now marketed world-wide. Thus the actual location of the server wasidentified using a DNS look-up tool.


Table 6 Adult Internet Responses by Ethnicity and Place of Response

reported place of response

Ethnicity CN

MI

CH

UU

K

GU

AM

KO

SRA

E

PALA

U

POH

NPE

I

RM

I

YA

P

HA

WA

II

USA

AU

ST

NE

W

EU

RO

PE

OT

HE

R

Carolinian 1Chamorro 1 1Chuukese 1 1Kosraean 1Marshallese 2 1 2 8 2 4Palauan 1American 1 1 3 1 13Australian 3 1 3New Zealander 1Other 1 1 1 2Total 2 10 1 1 1 5 1 24 5 1 1 6

Table 7. Origin and category of ISPs connecting to the initial page

Country Commercial Government Military Organisation School University Total %Australia 1 1 4 6 3.28Austria 1 1 0.55Canada 2 1 3 1.64Finland 4 4 2.19FSM 2 2 1.09Israel 6 6 3.28Italy 1 1 0.55Japan 4 1 5 2.73Netherlands 1 1 0.55New Zealand 1 1 2 1.09Palau 5 5 2.73United Kingdom 4 4 2.19USA 93 6 7 2 35 143 78.14Total 120 8 7 1 2 45 183 100.00


Table 8. Geographical Origin of ISPs in the USA connecting to the initial page compared to other parameters

On-Line QuestionnaireMicronesians

US Census 2000Marshall Islands Digital

Libraryn % n % n %

Alaska 359 0.50 133 0.12Alabama 5 3.7 502 0.70 632 0.57Arkansas 954 1.34 529 0.48Arizona 2 1.5 1759 2.47 1343 1.21California 10 7.4 22077 30.96 12539 11.34Colorado 43 31.9 1215 1.70 1702 1.54Connecticut 462 0.65 355 0.32Washington DC 3 2.2 471 0.43Delaware 251 0.23Florida 2177 3.05 6084 5.50Georgia 3 2.2 1531 2.15 1582 1.43Hawaii 13 9.6 8399 11.78 9804 8.87Iowa 1 0.7 245 0.34 438 0.40Idaho 283 0.40 894 0.81Illinois 918 1.29 2668 2.41Indiana 1 0.7 551 0.77 1177 1.06Kansas 376 0.53 619 0.56Kentucky 444 0.62 92 0.08Louisiana 558 0.78 529 0.48Massachusetts 3 2.2 484 0.68 1122 1.01Maryland 804 1.13 1195 1.08Maine 85 0.08Michigan 771 1.08 2251 2.04Minnesota 494 0.69 1979 1.79Missouri 1 0.7 1034 1.45 830 0.75Mississippi 92 0.08Montana 117 0.11North Carolina 1 0.7 1371 1.92 10352 9.36North Dakota 306 0.28Nebraska 1 0.7 305 0.43 238 0.22New Hampshire 112 0.10New Jersey 636 0.89 1499 1.36New Mexico 389 0.55 585 0.53Nevada 1 0.7 1377 1.93 544 0.49New York 1 0.7 1873 2.63 24442 22.11Ohio 2 1.5 712 1.00 2604 2.36Oklahoma 801 1.12 806 0.73Oregon 2 1.5 2840 3.98 2018 1.83Pennsylvania 4 3.0 876 1.23 2016 1.82Rhode Island 253 0.23South Carolina 482 0.68 717 0.65South Dakota 1 0.7 422 0.38Tennessee 713 1.00 818 0.74Texas 3 2.2 3971 5.57 5397 4.88Utah 572 0.80 995 0.90Virginia 4 3.0 1174 1.65 1536 1.39Vermont 32 0.03Washington 2 1.5 6278 8.80 2929 2.65Wisconsin 2 1.5 534 0.75 1710 1.55West Virginia 72 0.07Wyoming 1 0.7 57 0.05Korea (US Mil) 1 0.7Guam 16 11.9CNMI 1 0.7 4 0.00FSM 2 1.5 593 0.54TOTAL 130 100/00 71301 100.00 110547 100.00


In the case of US-based .com domains(51% of the sample), these were further brokendown by state. Based on the identified do-mains, the majority of the respondents camefrom the USA (41.4%). Only 19% of all re-sponses came from locations in Micronesia,with Guam (17.2%) providing the over-whelming majority. Thus the target audience ofthe questionnaire, Micronesians living in theDiaspora outside Micronesia was seeminglymet.

How does the origin of the people lookingup the on-line questionnaire compare to thegeneral population?

Table 8 compares the geographical origin ofpeople living in the USA looking up the initialpage; against (i) the distribution of Micro-nesians in the USA according to the census of2001; and (ii) the geographical origin of pagerequests to pages on the Marshalls IslandsDigital Library originating from US commer-cial providers for the period July 2002 toDecember 2003 (Spennemann 2004b). The re-sponses to the on-line questionnaire bear nocorrelation to the Micronesian population dis-tribution in the USA (r=0.16) or to the audi-ence of the Marshall Islands website (r=0.05).The correlation between the distribution of theMicronesian population and the audience ofthe Marshall Islands website is slightly better,but also not convincing (r=0.48).

So, what drives the response? At present wedo not know as the return sample is only small.Further research will be required.

CONCLUSIONSThe question arises, how effective an on-linequestionnaire is in a research setting in Micro-nesia. The questionnaire design was successfulin that it resulted in a quite respectable re-sponse rate, with a satisfactory overall comple-tion rate of all elements. Where the on-lineadministered questionnaire under performedwas in regard of the total number of question-naires submitted. That can be explained interms of the general paucity of internet con-nections in much of Micronesia as part of thewider digital divide issue (Spennemann 2004a).It is also a factor of limited promotion. Beyond

the initial posting on Yokwe.net and a fewnewsgroups, there was no avenue for con-tinued promotion of its availability beyondword of mouth. As a result, it seems, the useruniverse reached was only small.

The small population of responses meansthat the data collected in the survey are neitherrepresentative, nor can they be correlated andcompared with the data collected by traditionalpaper-based questionnaires administered to‘captive’ audiences. Future application of thetechnique in a Micronesian setting will requireadvance ‘advertising’ and the development of a‘survey culture’ among the Micronesian on-linecommunity.

BIBLIOGRAPHYAdams, W H, Ross, R E, Krause, E L & Spenne-

mann, D H R 1997, Marshall Islands Archaeology,Micronesian Resources Study, Micronesian En-dowment for Historic Preservation and U.S.National Park Service, San Francisco.

ASA 2000, Designing a Questionnaire, ASA Series -What is a Survey?, American Statistical Associa-tion.

Atrostic, BK, Bates, N, Burt, G & Silberstein, A2001, 'Nonresponse in U.S. GovernmentHousehold Surveys: Consistent Measures, Re-cent Trends, and New Insights', Journal of OfficialStatistics, vol. 17, no. 2, pp. 209-226.

Baron, J & Siepmann, M 1999, Techniques for creatingand using web questionnaires in research and teaching,University of Pennsylvania, viewed 10 May2002,<http://www.psych.upenn.edu/~baron/examples/baron4.htm>.

Bowker, D & Dillman, DA 2000, An ExperimentalEvaluation of Left and Right Oriented Screens for Webquestionnaires, Washington State University,Pullman, Washington, viewed 10 May 2002,<http://survey.sesrc.wsu.edu/dillman/papers/AAPORpaper00.pdf>.

Brady, S, Osborn, L, Olson, L & Blumberg, S 2003,'Questionnaire Design: Issues of language andCulture'.

Carucci, LM 1988, 'Small Fish in a Big Sea: Geo-graphical Dispersion and Sociopolitical Centrali-zation in the Marshall Islands', in State andSociety: Emergence and Development of Social Hierarchyand Political Centralization, J Gledhill & B Bender(Eds), Oxbow, Oxford, pp. 33-42.


Carucci, LM 1997, 'Irooj Ro Ad: Measures ofChiefly Ideology and Practice in the MarshallIslands', in Chiefs Today: Traditional Pacific Leader-ship and the Postcolonial State, GMLL White (Eds),Contemporary Issues in Asia and the Pacific,Stanford University Press, Stanford, California,pp. 197-210.

Carver , Steve; Kingston, Richard & Turton, Ian(1998) A Review of graphical environments onthe World Wide Web as a means of wideningpublic participation in social science researchhttp://www.ccg.leeds.ac.uk/agocg/report.htm

Christiansen, Henrik. 1994a. The Archaeology ofWorld War II in the Marshall Islands. Volume I:Tõrwa. Republic of the Marshall Islands: His-toric Preservation Office.

Christiansen, Henrik. 1994b. The Archaeology ofWorld War II in the Marshall Islands. VolumeII: Wõjjã. Republic of the Marshall Islands:Historic Preservation Office.

Christiansen, Henrik. 1994c. The Archaeology ofWorld War II in the Marshall Islands. VolumeIII: Jãlwõj. Republic of the Marshall Islands:Historic Preservation Office.

Christiansen, Henrik. 1994d. The Archaeology ofWorld War II in the Marshall Islands. VolumeIV: Mile. Republic of the Marshall Islands: His-toric Preservation Office.

Couper, MP, Traugott, MW & Lamias, MJ 2001,'Web survey design and administration', PublicOpinion Quarterly, vol. 65, no. 2, pp. 230-253.

de Vaus, D A 2002, Surveys in Social Research, 5thedn, Social Research Today, Allen & Unwin,Crows Nest, Australia.

Denfeld, D. Colt 1979a. Field Survey of Ponape:World War II Features. Micronesian ArchaeologicalSurvey Report No. 2. Saipan: Historic Preserva-tion Office, Trust Territory of the Pacific Is-lands.

Denfeld, D. Colt 1979b. Field Survey of Truk:World War II Features. Micronesian ArchaeologicalSurvey Report No. 6. Saipan: Historic Preserva-tion Office, Trust Territory of the PacificIslands.

Denfeld, D. Colt 1981. Japanese fortifications andother military structures in the Central Pacific.Micronesian Archaeological Survey Report No. 9. Sai-pan: Historic Preservation Office, Trust Terri-tory of the Pacific Islands.

Dillman, D A 2000, Mail and Internet Surveys: the tai-lored design method, 2nd edn, John Wiley & Sons,Inc., New York.

Dillman, DA, Phelps, G, Tortora, R, Swift, K, Ko-hrell & Berck, J 2001, Response rate and measure-

ment differences in mixed mode surveys, viewed 12August 2004<http://survey.sesrc.wsu.edu/dillman/papers.htm>

Dillman, DA, Redline, CD & Carley-Baxter, LR1998, The Web Questionnaire Challenge to Sur-vey Methodologists, viewed 10 May 2002,<http://survey.sesrc.wsu.edu/dillman/zuma_paper_dillman_bowker.pdf>.

Dillman, DA, Tortora, RD & Bowker, D 1998, Prin-ciples for Constructing Web Surveys, WashingtonState University, viewed 10 May 2002,<http://survey.sesrc.wsu.edu/dillman/papers/websurveyppr.pdf>.

Dillman, DA, Tortora, RD, Conradt, J & Bowker,D 1998, Influence of Plain VS. Fancy Design on Re-sponse Rates for Web Surveys., viewed 10 May 2002,<http://survey.sesrc.wsu.edu/dillman/papers/asa98ppr.pdf>.

DSS Research, I 2000, Complementary Methodolo-gies: Internet versus Mail Surveys, DSS Re-search, Inc, viewed 9 December 2002,<http://www.dssresearch.com/resources/PDF_Files/mail_v_internet.PDF>.

Flesch, R How to test readability, Harper and Brothers,New York.

Foddy, W 1993, Constructing Questions for Inter-views and Questionnaires. Theory and Practicein Social research, Cambridge University Press,Cambridge.

Frazer, L & Lawley, M 2000, Questionnaire design &administration, John Wiley & Sons Australia, Ltd,Brisbane, New York, Chichester, Weinheim,Singapore, Toronto.

Galvan, J 1998, 'Tangible Spanish Legacy in Micro-nesia', in Pacific Islands: The Spanish Legacy, J Gal-van (Ed), Ministry of Education and Culture,Spain, pp. 65-126.

Gillham, B 2000, Developing a Questionnaire, Con-tinuum, London and New York.

Gunn, H, Web Based Surveys, viewed 3 June 2002,<http://www.accesswave.ca/~hgunn/special/papers/websurv/index.html>.

Hager, M, Wilson, S, Pollak, T & Rooney, P nd, Re-sponse Rates for Mail Surveys of Nonprofit Or-ganizations: A Review and Empirical Test.

Harper, B, Slaughter, L & Norman, K, Question-naire administration via the WWW: A validation& reliability study for a user satisfaction ques-tionnaire, viewed 28 January 2003,<http://lap.umd.edu/webnet/paper.html>.

Hau'ofa, E 1998 'The Ocean in Us', The ContemporaryPacific: A Journal of Island Affairs, vol. 10, no. 2,pp. 392-410.


Hezel, F X & Graham, C 1997, Truk Underwater Ar-chaeology, Micronesian Resources Study, Micro-nesian Endowment for Historic Preservationand U.S. National Park Service, San Francisco.

Jobber, D 1989, 'An examination of the effects ofquestionnaire factors on response to an in-dustrial mail survey', International Journal ofResearch in Marketing, no. 6, pp. 129-140.

Joinson, A. N. (1999). Social Desirability, Ano-nymity and Internet-based questionnaires. Be-havior Research Methods, Instruments and Computers31 (3), pp. 433-438.

Joinson, A.N., Woodley, A., & Reips, U-R. (inpress). Personalization, authentication and self-disclosure in self-administered Internet surveys.In press, Computers in Human Behavior.

Kennedy, JM, Kuh, GD & Carini, R 2000, Web andMail Surveys: Preliminary Results of Compari-sons Based on a Large-Scale Project, viewed 15July 2002.<http://www.indiana.edu/~csr/AAPOR_NSSE.pdf>.

Mangione, TW 1995, Mail Surveys: Improving theQuality, Applied Social Research Methods Series,Sage Publications, Thousand Oaks, CA.

Moore, D S & McCabe, G P 1993, Introduction to thePractice of Statistics, 2nd edn, W. H. Freeman andCompany, New York.

O'Neill, JG & Spennemann, DHR 2002 'ColonialHeritage in post-colonial Micronesia', PacificStudies, vol. 25, no. 3.

O'Neill, Jon Graeme 2000, 'Management of Ger-man Colonial Heritage in the Pacific', CharlesSturt University, Albury, NSW, Australia.

Oppenheim, A N 1982, Questionnaire Design and At-titude Measurement, Heinemann Books on Sociol-ogy, Heinemann Educational Books Ltd,London.

Parker, P 1994, Protecting Historic Properties andCultural Traditions in the Freely AssociatedStates of Micronesia in Micronesian ResourcesStudy, United States Department of the Interior,National Park Service, Washington.

Ritter, Philip; Lorig, Kate; Laurent, Diana & Mat-thews (2004) Internet Versus Mailed Question-naires: A Randomized Comparison. Journal ofMedical Internet Research vol. 6 nº 3, pp.e29

Rolleston , Anna and Anderson, Helen, (2004) At-trition and retention. The voice of missing stu-dents. Australian Vocational Education AndTraining Research Association. The seventhAustralian VET Research Association Confer-ence "Learner & Practitioner - the Heart of theMatter" 17-19 March 2004 Rydges EaglehawkResort , Canberra . Paper PA 027.

<http:http://www.avetra.org.au/Conference_Archives/2004/documents/PA027Rolleston.PDF>

Ronckers, C, Land, C, Hayes, R, Verduijn, P &Leeuwen Van, F 2004, 'Factors ImpactingQuestionnaire Response in a Dutch Retrospec-tive Cohort Study', Annals of Epidemiology, vol.14, no. 1, pp. 66-72.

Saphira M.; Glover M. (2000) New ZealandNational Lesbian Health Survey. Journal of theGay and Lesbian Medical Association, June 2000,vol. 4, no. 2, pp. 49-56.

Schonlau, M, Fricker Jr, RD & Elliott, MN 2001,Conducting Research Surveys via E-mail and the Web,RAND, viewed 9 December 2002,<http://www.rand.org/publications/MR/MR1480/>.

Sefton, P & Atkinson, J, (2004) Can an online poll beused as a valid alternative to a traditional paper basedsurvey?, viewed 2 July 2004<http://ausweb.scu.edu.au/aw04/papers/refereed/sefton/paper.html>.

Spennemann, Dirk H.R. (1992a) World War II Re-mains on Central Pacific Islands: Perceptions ofHeritage versus Priorities of Preservation. ThePacific Review 5 (3), 1992, 278-290.

Spennemann, Dirk H.R. (1992b) Apocalypsenow?–the fate of World War II sites on theCentral Pacific Islands. Cultural Resources Man-agement [U.S.National Park Service, Washington]15(2), 1992, 15-16, 22.

Spennemann, Dirk H.R. (1993) Ennaanin Etto - ACollection of Essays on the Marshallese Past, HistoricPreservation Office, Majuro, Republic of theMarshall Islands.

Spennemann, Dirk H.R. (2003) 'Teacher and Stu-dent Perceptions of the Cultural Heritage of theCNMI', Micronesian. Journal of the Humanities andSocial Sciences, vol. 2, no. 1-2, pp. 50-57.

Spennemann, Dirk H.R. (2004a) Digital Divides inthe Pacific Islands. IT & Society vol. 1 nº 7, pp.43-63.

Spennemann, Dirk H.R. (2004b) A Digital libraryand archive about the Marshall Islands: experi-ences and challenges. Australian Library Journalvol. 53, nº 3, p. 235–256.

Spennemann, Dirk H.R. & Meyenn, Robert J.(1997) Cultural Heritage management and Cur-riculum Development: Some issues facingMelanesian Nations. Johnstone Centre of Parks,Recreation and Heritage Report. 98. Albury,NSW: Charles Sturt University.

Tun, P 1996, 'Dances, Chants and Songs as YapeseArt Forms' in 5th Pacific Islands Association ofLibraries and Archives Conference (PIALA '95):


Preservation of Culture Through Archives andLibraries, Ed A Cohen.

Ward, Wesley & Spennemann, Dirk H.R. (2000a).Meeting local needs? a study of a communica-tion project established in the Pacific Islands.Public Administration and Development 20(3), pp.185–195.

Ward, Wesley & Spennemann, Dirk H.R. (2000b).Getting wired: a Pacific Islands study. AustralianJournal of Communications 27(3), pp. 91-105.

Ward, Wesley Stewart 1997, 'A study of reasonswhy people in the Pacific Islands connected tothe Internet', Charles Sturt University, Albury.

WebSurveyor Corporation 2000, Survey Design Con-siderations, WebSurveyor Corporation, viewedMay 2002, <http://www.websurveyor.com>.

WebSurveyor Corporation 2000, Survey Design Con-siderations, WebSurveyor Corporation, viewedMay 2002, <http://www.websurveyor.com>.

Wendt, A 1976 'Towards a New Oceania', Mana Re-view, vol. 1, no. 49-60.

Wham C.A. & Worsley A. (2003) New Zealanders'attitudes to milk: implications for public health.Public Health Nutrition 6 (1), pp. 73-78.

Zwimpfer Communications (1999) Connectednessin Pacific Islands Countries. A survey on the useof computers, e-mail and the Internet in educa-tion, culture and communication. Apia:UNESCO Office for the Pacific States<http://www.unesco.org/webworld/publications/99_Internet_survey_report.rtf>

Zwimpfer Communications (2002) Internet Infra-structure and e-Governance in Pacific IslandsCountries. A Survey on the Development andUse of the Internet. Apia: UNESCO Office forthe Pacific States.<http://www.unesco.org/webworld/publications/2002_Internet_survey_report.rtf>

AUTHOR BIOGRAPHIES AND CONTACTJon O’Neill has an Applied Science Degree inParks, Recreation and Heritage with First ClassHonours. He has conducted research in severalMicronesian political entities including Guam,Commonwealth of the Northern Mariana Islands,Federated States of Micronesia and the Republic ofthe Marshall Islands. He is a doctoral student atCharles Sturt University where he is presently re-searching heritage issues in Micronesia. Particularattention is being given to: changes in preservationmanagement that may have occurred following self-government, evolving Micronesian perceptions ofheritage values and the extent to which indigenousdecision-makers have been empowered.

Jon O’Neill, Charles Sturt University, P.O.Box 789, Al-bury NSW 2640, Australia e-mail [email protected]

Dirk HR Spennemann is Associate Professor inCultural Heritage Management at Charles SturtUniversity, Albury, Australia. His main research in-terests are German colonial heritage in Oceania, inparticular Micronesia, and historic preservation is-sues in Micronesia in general. His second focus rethreats to heritage posed by natural and human haz-ards and threats posed by managers in their effortsto counter these hazards. Ethical Heritage Planningand Policy are the cornerstones that need to be un-derstood and addressed if our past is to have ameaningful future.A/Professor Dirk H.R. Spennemann, Charles Sturt Uni-versity, P.O.Box 789, Albury NSW 2640, Australia [email protected]

micronesian - charles sturt universitymicronesia.csu.edu.au/mjhss/issue2004/mjhss2004_06.pdf ·...

Documents