Methodology

Home

Abstract
Introduction and Research Question
Methodology
Findings and Literature Review
Statistical Charts
Analysis
Conclusions and Recommendations
Works Cited
Links
An Evaluation of the United Nations Volunteer Program in Kosovo

Kosovo UNV Job Satisfaction Questionnaire

Kosovo is a distant place, distant for some geographically, but for nearly all in terms of contact with the world beyond its frontiers. With newly repaired but spotty telephone networks, satellite-driven e-mail in many cases, and new - but for the requirements of my research, unreliable - postal service, distant field research posed significant challenges (1). However, these challenges are effectively met with a self-administered online questionnaire.

This research instrument demonstrated many advantages:

- Allowed respondents to conveniently reply on their own time

- Maintained discretion in the work setting

- Provided the option of anonymity, while assuring confidentiality

- Extremely cost effective (virtually free)

- Facilitated high response rate

- User friendly

- Increased candor in responses

Additionally, the Internet address (or URL) of the questionnaire was included as a hyperlink within the solicitation, which greatly increased the ease of use. Rather than cutting and pasting an attached questionnaire, filling out the answers, and e-mailing it back as another attachment, the participant needed only to click on the URL, which would then bring the questionnaire up as a separate web page. Once at this website, the participant could click radio buttons to respond to most questions, type details or clarifications in the open-ended response fields, and then click the submit button at the bottom of the page.

These advantages heavily outweighed the few potential drawbacks that might be considered. The validity of responses, for example, might be called into question because of the unsecured nature of online communication. While I gave some consideration to the potential for falsified responses, I went forward with the knowledge that this is a possibility with any questionnaire, online or otherwise. Electronic mail accounts tend to be closely guarded by their subscribers (Dyson, 1993), and therefore the likelihood of another person getting into a UNV account simply to send an invalid response seemed unlikely. The other possibility that a respondent would fill out completely false information cannot be discounted entirely, but the personal nature of the questionnaire and the level of realistic representation in the responses suggest that this was not the case.

Also of concern during the first round of questionnaires was the potential for sampling difficulties while using the snowball technique. This technique is a way of broadening sample size by asking participants to forward the request to others who may be interested. This method can be useful, but adds an unknown factor to the research; it is not always certain who receives such second-hand requests, or if they are appropriate candidates for the survey. However, with an initial fear of insufficient sample size, I decided to let the snowball roll anyway, despite the uncertainty.

I found that the online questionnaire was extremely effective for the reasons bulleted above, though I did encounter certain difficulties with its dissemination. First was the question of how and to whom to send the request for participation. I decided that all Kosovo UNVs, past and present, would be useful to my research, especially since the Kosovo Mission had by that point had a short year-and-a-half lifespan. Systematic contact with these UNVs was another issue, as finding e-mail addresses through the UNV support offices was rather difficult (NB- see below for notes on the second round of questionnaire responses, with the help of a comprehensive UNV e-mail list). When I initially received a list of UNVs from the UNMIK (Civil Administration) UNV Support Office, it contained 188 names but only 25 e-mail addresses. Some of these were next of kin contact information and therefore unreliable, and others proved to be dead addresses. I also received a much shorter list from the United Nations Development Program (UNDP) UNV Support Office (2), which only provided nine valid addresses.

I had anticipated such difficulties before I departed the province, and had made an effort to compile my own list of e-mail addresses for my imminent research. I asked all UNVs I knew for their e-mail addresses, and for any other e-mail addresses of UNV colleagues outside of my immediate sphere. While clearly a subjective and non-random selection method, I acted to prevent a complete dearth of potential respondents. The address list I compiled on my own totaled twenty, four of which later proved to be dead.

The discovery of dead or mistaken addresses came about after the first step of sending out a pre-invitation to participate in early December 2000. This technique is seen as proper online research etiquette, and was issued to notify participants in advance and create a sense of participation commitment (Thach, 1995). Inactive addresses were either blocked by Hotmail (my e-mail provider), or later by other servers. One recipient of the pre-invitation wrote back that, though unsure of how I received his address, he was sure of never having been a UNV, and asked that I please delete his name from my list(3).

After many draft versions and revisions of the online questionnaire, the final version was completed in late January, nearly a month and a half after the pre-invitation. Anxious because of this time lapse, I somewhat hastily sent an e-mail notification to potential respondents, with the address of the questionnaire website within the body of the message. My haste, Hotmails temperament, or some combination of the two prevented a successful launch in that first attempt. I included my e-mail address in the list as a test, and quickly realized that my first solicitation to participate had failed. I quickly sent out another request, this time to individual addresses rather than large groups (suspecting that the bulk mailing might have been the culprit). This time, as far as my test showed, the launch was successful.

It appeared that Hotmail began to immediately experience widespread network problems soon thereafter, as I was unable to log on to my account on and off for several days. This experience, and the awareness that Internet access in Kosovo could be unreliable, left me unsure if I should wait for responses, or send out a third attempt.

Though I did receive a few responses in the next few days, I decided after a four-day wait to send yet another message. This time I sent a reminder to respondents under the pretext of uncertainty from the recent Hotmail difficulties. This one seemed to hit the mark, and a large portion of my responses arrived within the following week.

After the initial wave of responses, I continued completed questionnaires until late February 2001. The lag time might be explained as a result of the snowball technique, described above, where participants pass on my request and questionnaire web address to other Kosovo UNVs (as I had indicated in the postscript of the original requests), who may take longer to respond than those originally solicited.

From the total seventy-five requests sent out in January 2001 (with at least eighteen dead addresses), I received thirty-one responses.

Second Round

I would have relied solely on the first batch of responses as a satisfactory foundation for my research project, had an additional opportunity not presented itself. I was asked by my former supervisor in UNMIK to return to Kosovo in the spring of 2001, this time as a consultant for the shelter program I had previously managed, and without affiliation to UNV. By that time, a comprehensive e-mail address list of all UNVs remaining in the mission had been compiled, and I was able to obtain a copy of it within days of my arrival. Enabled with an objective list, as originally desired in late fall 2000, I was now able to test the validity of the first round of findings, broaden the research sample, and offer many more Kosovo UNVs the opportunity to share their feelings about the program.

With the experience from the first round of questionnaires, I made sure to test run the second solicitation, which resulted in a successful launch on the first attempt in early March 2001. This time, I sent requests only to those addresses that were not included in the first run. To eliminate the inherent uncertainty when using the snowball technique, I also removed the postscript, which had earlier asked participants to pass the message along to other UNVs. This request, sent to 222 addresses, resulted in twenty-four responses within days of its mailing.

In total, I sent requests for participation to 298 UN Volunteer e-mail addresses, and from these at least thirty-three proved to be inactive or incorrect. From the maximum of 265 remaining addresses, I received a total of fifty-four valid responses, for a response rate of 20.4%.

Findings and Literature Review

1) Conducted from Brattleboro, Vermont, in the winter of 2000-2001.
2) The UNDP UNV Support Office administers to UNVs within the agencies not directly working with the United Nations Mission in Kosovo Civil Administration, e.g. UNHCR. The UNMIK UNV Support Office is responsible for the bulk of UNVs, who serve directly within Civil Administration. Both offices report to UNDP, and have the same guidelines for support and training of UN Volunteers.
3) This address had been provided by the UNV support office.