The Asia Foundation, with support from the USAID Democracy and Human Rights Program, has funded programs at the Cambodian Institute for Human Rights (CIHR), which are aimed to train government officials and others in the areas of democratic governance and elections.
CIHR has been implementing two programs within this framework. One, the Good Governance/National Election Education Program (GG/NEE) was aimed at provincial leaders and was designed to raise the level of understanding and support for free and fair elections. The other, Preparing Local Leaders for Commune Elections (PLLCE), was aimed at commune chiefs and other local leaders and was likewise designed to stress the importance of elections in a democracy and the need for free and fair elections.
The Asia Foundation requested that an independent, systematic evaluation be undertaken to assess the effectiveness of these two training programs. TAF and USAID and CIHR all desired reliable and accurate information regarding the influence of the training courses on the knowledge and attitudes of government officials and others who participated in the courses.
The evaluation results are likely to be helpful to CIHR in making improvements in current programs and in creating systematic internal evaluation procedures. The results are also likely to be useful to TAF and USAID in their decision making regarding future programs and in their determination of the types, scope and coverage of public education in the area of democracy and human rights that should be supported once the elections of July 1998 are over.
The program proposal describes the aims of CIHR to reach key government officials at the provincial level who are considered to be most critical to the success of free and fair elections. These key officials were intended to be the audience for a 5-½ day educational program known as the gMajor Seminar.h In addition, simpler, shorter seminars were intended to reach additional officials. These shorter presentations were known as gMini Seminars.h
According to the proposal, the subject matter of both Major Seminars and Mini Seminars was designed to center on the role of elections in a free society and on the proper function of officials in ensuring the successful implementation of free and fair elections. The intended theme of both the seminar series was that officials must be impartial in the election process.
According to the proposal, the core message of the training was intended to be that the democratic system is the best means of protecting everybodyfs rights. The design of the seminars was intended to persuade government officials that in a democratic system their rights and jobs would be protected, and that, as career civil servants, a change in government need not mean a loss of their job or other worse disadvantage to them. The appeal of civil servant job security under a democratic system was intended to generate the cooperation of these officials in a free and fair electoral process.
The Center for Advanced Study (CAS) was selected to conduct the evaluation because of its familiarity with survey research techniques and because of its research track record in the provinces on topics related to democracy and voter education.
At the beginning of June 1998, CAS, with the cooperation of CIHR, assembled the Khmer course materials used in the training in both the GG/NEE and PLLCE programs. There were differences in the implementation of the curriculum in the two programs. However, we determined that the content of the courses was similar enough that a single questionnaire for participants in both programs would permit a satisfactory assessment of the retention and understanding of the curriculum materials. The questionnaire tested the utilization by participants of CIHR media supplements to the training courses. The questionnaire also included several open-ended questions that probed the perceptions of the participants regarding changes in attitude or behavior that they could attribute to the CIHR training.
CAS researchers used the Khmer language curriculum materials of the CIHR training to construct a survey instrument. An English translation was made, fine tuned and a draft of the questionnaire was pre-tested. Respondents could complete the final streamlined and tested version of the questionnaire in about 45 minutes.
A sample of approximately 10% of training course participants during the period mid-June to mid-July 1998. CIHR provided us attendance lists for the training sessions, from which we could draw our samples. The lists of trainees were organized by provincial venues where the courses were held. We used those provinces as our initial sampling frames for each of the two programs.
For the GG/NEE population, we had to exclude Ratanakiri and Siem Reap from our sampling because of limitations in the budget. Otherwise, we drew a random sample from each provincial center where CIHR had conducted its training courses. For the PLLCE population in four target provinces, random participants were chosen from a sampling frame of randomly selected districts within each of these provinces.
Within the GG/NEE population, CIHR addressed its training to two sub-groups, which we sampled separately. One group received a 5-½ day course that CIHR called the gMajor Seminar.h The other group, called the gMini Seminar,h received training in modules scattered over time. The members of this group could have received as little as one day or as much as five days of training, which, on average, would expectably depress the scores of this group, compared with the Major Seminar participants.
In some provincial offices we experienced resistance to our research, and our team was only able to elicit cooperation with the survey by referring to the letter from the Ministry of Interior authorizing the conduct of the research. But our researchers noted that this invocation of higher authority might also have had the effect of occasionally frightening our designated sample respondents away. Apparently when government office workers overheard the CAS team explaining their intention to conduct the survey on the basis of randomly selected individuals named on our list, the officials named were found to have discretely left their offices, making themselves unavailable to participate.
The ethos of government service that has prevailed in Cambodia in recent years is probably at the root of the fearfulness of officials, whose names appear to be on a Ministry of Interior list, to be interrogated by unknown visitors. This is simply one of the conditions any survey among Cambodian government officials has to face.
In its original preparation for the PLLCE program, CIHR undertook a baseline survey of attitudes among commune chiefs and village notables in the four target provinces of the Northwest where the CIHR training was to be targeted. A gcontrol grouph of commune chiefs and notables in provinces that had not been targeted for training was also surveyed.
The terms of reference of this evaluation followed the original CIHR design in calling for a survey of a control group of respondents at the commune level who had not participated in the CIHR training for commune leaders, PLLCE.
Our original intention was to survey commune leaders and village notables in the same control provinces CIHR had originally surveyed, in order to have a comparison with the scores achieved by the commune leaders who had been trained.
This turned out not to be possible. Our letters of introduction from the Ministry of Interior, which were sent to the relevant Governorsf offices, were only sent to the provinces where CIHR had conducted their training programs. Of the three gcontrol grouph provinces that CIHR had originally surveyed, one was also the site of GG/NEE training. So there would have been no problem about conducting our survey in communes of that province. However, the other two provinces had not received authorization from the Ministry of Interior to allow this kind of local survey so they denied permission for our team to conduct the research in the communes of these provinces.
Nevertheless, in the provinces where no CIHR training had been conducted, the authorities were more or less reluctantly agreeable to allowing our researchers to conduct the survey in an informal and voluntary basis in their provincial offices. This became our gcontrol group.h Our Control Group consists of government officials at the provincial level comparable to those who had received the GG/NEE training, accordingly, our Control Group findings should be compared primarily to the Major Seminar and Mini Seminar participants sampled. However, comparisons of all four groups with one another, the Major Seminar and Mini Seminar groups of GG/NEE, the Control Group and the PLLCE, is also instructive.
Our best indication of CIHR impact on its trainees will come from a comparison of achievement in each knowledge area between the Major Seminar component of the GG/NEE group and the Control Group. Both these groups consist of middle to high-ranking provincial level government officials. The Major Seminar participants received the full five and one half day CIHR training. The Control Group received no CIHR training. The Control Group reports rather little formal training of any kind.
The increase in scores of the trained group over the untrained group, it can be argued, is due to the training. The greater the difference in achievement in a particular knowledge area between the groups would indicate the portion of the curriculum where the CIHR impact was more significant.
These comparisons of achievement are based on the summary of scores for the four groups surveyed, which follows in graphic form in the following section (E). These summaries, in turn, are based on the detailed findings for each group, which are presented below in this report under the headings: II. The Good Governance/National Election Education Program,(GG/NEE) III.Control Group, and III. The Preparing Local Leaders for Commune Elections Project (PLLCE).
The greatest impact of the CIHR courses is apparent in the core course areas "Good Governance" and "Democracy and Elections."
In "Good Governance," the average score achieved (the mean score) increased 6%, from an average of 89% among the Control Group to an average of 94% in the Major Seminar group. The impact of the CIHR training is much more evident, however, in the increase of perfect scores in this subject area. The rate of perfect scores in the Control Group was 33%, and in the Major Seminar group it was 59%. This is an increase of 79% in mastery of the subject matter, which is very likely due to the training. The rate of high scores shows the same pattern. The rate of scores ³ 93% correct shows an increase of 29% in the Major Seminar group, over the Control Group.
In "Democracy and Elections" the same dramatic increase in achievement is apparent in the trained group. The average score in the Control Group is 94% correct responses, while the average in the Major Seminar group is 98%, an increase of about 4%. But when we look at the achievement of perfect scores, we see a rate of 43% in the Control Group and a rate of 81% in the Major Seminar group. This is an increase in mastery of 88%, due to the training. The rate of high scores, ³95% correct, shows an increase of 24% in the Major Seminar over the Control Group.
Somewhat less dramatic results, but still highly creditable evidence of the CIHR training impact are apparent in the subject areas "Human Rights" and "The Constitution."
In both these subject areas, the increase in mean score of the Major Seminar over the Control group is in the range of 1-2%. However, much bigger gains are visible in the increased rate of perfect scores. In "Human Rights," the rate of perfect scores in the Control Group is 38%, and in the Major Seminar group it is 54%, indicating an increase in mastery of 42%, due to the training. In "The Constitution," the rate of perfect scores in the Control Group is 32% and in the Major Seminar it rises to 43%, indicating an increase of 34%, attributable to the training.
In the subject area of "The Constitution" we may note that the Commune Leader group performs particularly poorly, with a rate of perfect scores at 14%. This is a subject area that is probably not discussed in as much detail in the PLLCE curriculum as in the GG/NEE curriculum.
In the two subject areas "Conflict Resolution" and "Development" there is a healthy increase in scores due to training, but at a relatively slighter level than registered in other subject areas.
The increase in average score in these two subject areas of the Major Seminar over the Control Group is in the range of 1-2%. In "Conflict Resolution," the Control Group achieved a rate of 79% of perfect scores, and the Major Seminar shows a rate of 88% perfect scores. This is an increase of 11% due to the training. In "Development," the Control Group shows a rate of 56% perfect scores, and the Major Seminar shows a rate of 65%. This is an increase of 16% due to the training.
In the subject area "Gender Awareness" the training seems to have raised the average of the Major Seminar group 6% over the average for the Control group. Remarkably, the average score for the Commune Leader group is greater in this subject area than the averages for the other groups. This is the only occasion when a PLLCE average score exceeds a GG/NEE average score.
In the subject area "Management" there is little difference in the average score or in the rate of perfect or high scores among the groups surveyed.
In Media Impact, the score of the Control Group was 50% and that of the Major Seminar group was 68%. This is an increase of 18% presumably due to the emphasis placed on the concurrent media presentations for the trainees. The Commune Leader rate was even higher, at 82%. This is probably more due to the mandated duty of village leaders to listen to the radio news than to the CIHR training.
In the area of participant perceptions of the effectiveness of the CIHR training, we see great similarities between the Major and Mini Seminar participants of the GG/NEE training, which contrast somewhat with the findings from the Commune Leader participants in the PLLCE training.
The GG/NEE participants indicate that training in "Management," "Rule of Law," and "Consultation in decision-making," in that order, had the greatest impact on them and their colleagues. This result indicating a strong perception of the importance of "Management" seems incongruous in light of the finding that the CIHR training seemed to have little effect on achievement in "Management" as a knowledge area.
The explanation is probably that the CIHR curriculum presented "Management" less as a body of knowledge than as a practice. The practice of good management was evidently linked by CIHR training to the principles of good governance and democratic consultation, as they apply to the workplace. The perception of the importance of "Management" seems to reflect a perceived improvement in working conditions for the provincial officials due to CIHR training.
The GG/NEE participants held that "Impartiality" of government officials was the most significant factor they now recognized as an attribute of good leadership. "Election fraud" was registered as the most serious concern for leadership. This finding confirms the success of the CIHR training in meeting its objective to stress impartiality of government officials in its program.
The PLLCE participants held that training in "Rule of Law" was the most significant area they and their colleagues had been exposed to in the training, followed by "Management" and "Consultative decision-making." As with the GG/NEE participants, the Commune Leaders registered "Impartiality" as the most significant hallmark of good leadership, while "Corruption" and "Electoral fraud" were nearly equal in characterizing bad leadership.
The message of impartiality of government officials was stated as central to both of the CIHR training programs. This message was apparently very successfully conveyed to the participants whom we sampled in this survey evaluation.
***
The most striking difference between the Commune Leader group and the Control Group is in the scores for Human Rights and for The Constitution. The latter subject may be too technical or legalistic for Commune Leaders, but an acquaintance with Human Rights should surely be a central concern for these local level leaders.
***
|
Mean Scores |
Major Seminar |
Control Group |
|
1.Good Governance |
94 |
89 |
|
2. Democracy and Elections |
98 |
94 |
|
3. Management |
97 |
96 |
|
4. Human Rights |
95 |
94 |
|
5. Constitution |
91 |
89 |
|
6. Conflict Resolution |
96 |
95 |
|
7. Development |
86 |
84 |
|
8. Gender Awareness |
96 |
90 |
|
9. Media Awareness |
68 |
50 |
|
Perfect Scores |
Major Seminar % of respondents |
Control Group % of respondents |
|
1.Good Governance |
59 |
33 |
|
2. Democracy and Elections |
81 |
43 |
|
3. Management |
94 |
93 |
|
4. Human Rights |
54 |
38 |
|
5. Constitution |
43 |
32 |
|
6. Conflict Resolution |
88 |
79 |
|
7. Development |
65 |
56 |
|
8. Gender Awareness |
90 |
90 |
|
Demographic
Variables |
GG/NEE |
PLLCE |
Control |
|
|
|
Major. |
Mini |
|
|
|
Gender m/f (%) |
90/10 |
72/28 |
94/6 |
73/27 |
|
Age (mean years) |
44.2 |
37.1 |
47.3 |
35.9 |
|
Schooling level (mean years) |
11.6 |
10.7 |
7.8 |
9.7 |
|
Years in Government Service (mean) |
18.6 |
14.2 |
16.4 |
12.9 |
|
Years in Present Position (mean) |
14 |
11.8 |
15.1 |
10.6 |
|
Months since CIHR course (mean) |
6.3 |
6.6 |
4.1 |
- |
|
Participation in other non-CIHR courses (%) |
18 |
22 |
27 |
4 |
The responses to the open questions 102 and 104 make it clear that for both Major and Mini Seminar participants, the main message they got regarding their own effectiveness was that working conditions for officials should be improved. Under the rubric gmanagementh the respondents typically praised the CIHR training course for introducing rational planning concepts and human resource administration values to them and their colleagues.
Likewise, under gconsultationh the principal concern of the officials was that discussion take place among all affected officials, rather than having unilateral decisions made by the head of the office.
gRule of Law,h gImpartialityh and gGood Governanceh responses were divided between those who applied these notions to government officials in their dealings with one another, and those who thought that these ideas applied in relations between government officials and the governed.
The responses to open question 108 make very clear that the CIHR message that government officials should be impartial in their attitude toward the elections was effectively transmitted to the respondents we were able to access. At the same time, the difficulty we had in locating our original sample of randomly chosen trainees was due in large part to the partisan election activities of many of the CIHR training course participants during working hours.
The responses to open question 109 indicate that the problems of election fraud and unfair treatment of opposition parties were the main examples of bad governance that they were willing to describe.
1. Background data for every participant should be collected in order to maintain a profile of the kinds of trainees that are benefiting from CIHR training. This data may also help fine tune the curriculum in light of the trends in needs of different kinds of officials.
2. A pre-test and post-test and various assessments during the course would enable closer monitoring of curriculum understanding and variation within the class during the training period. The assessments should be returned to the trainees, after having been noted by CIHR, so that the trainees have a convenient list of the highpoints of the lessons, and a record of their own improvement. It would be very easy for the CIHR trainers to describe the general tendencies or averages of any assessment so the trainees would know how they stood in relation to others, but with anonymity preserved.
3.
CIHR should begin its own systematic evaluation of its effectiveness in making
an impact on the understanding of the participants in its training sessions,
perhaps using the present evaluation as a model.
4. The CIHR Media Utilization might be improved by presenting some clips during the training or exercises so the participants could recognize the material when it appeared on the television or radio. Trainee responses to the media presentations during the training session would be useful, but would depend on careful coordination of scheduling.
5. The rationale of an appeal to job security for civil servants in a democratic system, regardless of election results, does not seem so persuasive, now that the elections have passed. In the next phase of training, an alternative appeal to prospective government officer trainees for the utility of democratic governance may have to be formulated, aside from improved working conditions for civil servants in a democratically governed office setting.
6. It may be worth considering how to broaden the scope of the training to generate a demand for good governance by the public, in addition to the training that focuses on improving the supply of good governance from officials.
1. In view of the kinds of scores that can be achieved without training, by a group with high motivation, a thorough analysis of a prospective training course audience by way of pre-tests is strongly indicated.
2. The use of monetary incentives to participation in the training has the potential to feed the traditional patronage mechanism and fill the training course with less motivated students and officials who may have served long in authoritarian regimes. This may be highly desirable in that the training would then be likely to reach the participants most in need of the course.
3. At the same time, intensified and attractive media coverage on a nation-wide basis of the subject matter of the training course might have the effect of increasing general media awareness and serving the individuals who, by their own efforts, already possess a high level of knowledge in the subject areas.
1. Human Rights and the basic guarantees stated in The Cambodian Constitution should be reinforced in the Commune Leader training program to bring the scores up in this group.
2. The extensive use of radio by the Commune Leaders to keep informed suggests a very cost effective means of reaching this group, and the village public, with distance learning on the topics of a good governance and democracy training course. TV is a less effective means to reach most Cambodian villagers, although it may be a more successful way to reach the rich elite of the provinces.
3. If preparation for future commune chief elections is the goal of these CIHR training exercises, then some consideration should be given to how civil society actors who are not necessarily presently connected to the state, can emerge as contestants in those elections. A strategy sensitive to the grassroots cultural and political traditions of Cambodian villages should be devised to support a voter education campaign that could be linked to Commune Leader training.
That campaign would aim to strengthen the awareness of the public of its right to demand good governance and accountability from its locally elected officials. The campaign would also emphasize that these commune elections are an opportunity to make a real choice about local leadership, rather than simply making a choice between/among candidates pre-selected by the state.
Such a voter education campaign could well parallel such projects as this PLLCE program aimed at government officials, or could be incorporated into an expanded PLLCE concept.
In a number of instances, the respondents asked for a blank copy of the survey because they found it a convenient summary of the highpoints of the training course. This suggests a recommendation that periodic assessments, like the one we conducted, should be integrated into the training course. Once the assessments were tabulated by CIHR for their evaluation purposes, the assessment papers could be returned to the trainees, so that they would have a record of the subjects covered and an indication of their improvement over the assessment period.
I. Methodology
A. Survey of Trainees
The CAS research team was led by experienced educators. The portion of the questionnaire that aimed to assess the knowledge of the participants was designed to be a kind of final examination based on the Khmer curriculum materials used by the CIHR trainers. These questions all asked for yes/no responses, and were organized according to core subject areas covered in the training.
By mid-June, the Khmer draft questionnaire was translated into English, and fine-tuned, again in consultation with the CIHR trainers and advisers. The instrument was then subjected to a gpre-test.h The pre-test was conducted among staff of several human rights NGOs, among the staff in the Ministry of Interior responsible for training and education, with whom CIHR regularly works, and among students and teachers at the University of Phnom Penh.
The pre-test enabled us to refine the Khmer language of the instrument, while remaining within the discourse and the vocabulary specific to the CIHR curriculum materials. The instrument was also trimmed to a size that could be completed in about 45 minutes.
We followed a stratified random sampling procedure, as described in the Report.
Naturally there were cases when a person designated to be included in the sample could not be reached. In that case, an alternative name was chosen at random from within the same provincial list or sampling frame. In Takeo, Kandal, Phnom Penh, Prey Veng and Pursat, it was particularly difficult to find people initially picked in our random sample. The explanation our researchers received was that government pay was too low so many officials had to have another job elsewhere, so did not appear at their office.
Since the pay scales are similar in all provinces, this explanation probably really reflects a combination of higher cost of living in these areas and a tacit understanding in these areas that gmoonlightingh was permissible.
Another factor affecting the probabilistic character of our random samples, especially in the later portion of our survey period, was the approach of the National Elections scheduled for the last week of July. Many of the people on our lists of CIHR training course participants were simply not available in their government office because they were engaged in partisan political activity during that time.
It must be acknowledged that, to some extent, the presence of a CIHR trainee in his or her office at this time, to answer our survey questions, is an indication that the message of government official impartiality in the elections had reached these respondents. But this introduces a bias into the sample that probably tends to inflate the scores over what they would have been if the survey had been conducted at another time, when those trainees who apparently had not adopted the core training message of impartiality might have been available.
The demographic profiles of the two sub-groups of GG/NEE are very different, adding a further complication to comparisons between them. Our researchers learned that mainly the close friends and favorites of the local coordinator of the training (usually a Deputy Director of Cabinet in the Governorfs office) were selected to participate in the Major Seminar training. There was evidently an incentive in the form of per diem for the 5 ½ day training that was perceived to be a resource to be shared especially with friends and fellow party members.
In contrast, participation in the Mini Seminar was evidently based more on personal interest and initiative than on a monetary reward. That element of self-selection to the Mini Seminar might be part of the reason for the finding that scores in this group are nearly as high as those in the Major Seminar, even though their exposure to the curriculum was less complete.
Another factor to mention in describing the implementation of the survey was the general reluctance among GG/NEE provincial center CIHR trainees to participate in this survey. This was in marked contrast to the general willingness, and even eagerness, of the PLLCE commune leaders to cooperate in answering the questionnaire.
In many instances our persistent researchers were finally able to complete the surveys, especially with the higher officials who were among the designated sample. But this often required many days of return visits early in the morning to the office, and visits to the officialsf homes. The effect on our research was mainly that the survey took longer to complete than we had contemplated, and that pushed us even closer to the late July elections.
In administering the survey to the sample of participants in the CIHR training, our researchers often encountered what should probably be called gexam anxiety.h These respondents were very worried that they might give the wrong answers, or that somehow they, or their group, might be identified as falling below an acceptable average. One of the ways the respondents dealt with this anxiety was to have our team leave the questionnaires overnight with the Deputy Chief of Cabinet, to be distributed by him to the designated sample of participants. Another explanation for this procedure might be that the Deputy Chief of Cabinet, who had originally coordinated the seminars with CIHR, wanted to see the responses of his officials. This breach of anonymity was unfortunately out of our control. In any case, this is a possible source for inflation of scores, as the respondents had a chance to review their training materials.
Another stratagem the exam-anxious respondents used was to ask their office colleagues (who had presumably also taken the training but were not selected in our sample) what they thought the correct answer was. Our young researchers were not in a position to contradict or oppose these attempts to improve the participantsf test results. But it is another possible source of inflation of scores in the GG/NEE sample.
The PLLCE project proposal describes the aims of CIHR to reach commune leaders in the four northwestern provinces in order to ensure that these officials understand and will fulfill their roles as elective officials in a democracy. The proposal notes that these grassroots officials will have a key role to play in assuring that the scheduled elections will be free and fair and without violence.
The goals of this project follow the goals formulated for the GG/NEE program quite closely. The training is intended to stress the theme that impartiality of local officials will be a key to preventing election-related violence and in assuring acceptance of election results.
The curriculum for the training in the PLLCE project is, as far as we could ascertain from the materials we obtained from CIHR, quite comparable to the curriculum presented to the Major and Mini Seminars of the GG/NEE project. However, there were no doubt differences in the delivery of the curriculum in view of the differences in level of sophistication between village commune leaders and provincial officials in the towns.
For purposes of this evaluation, we utilized the same survey instrument with both the PLLCE population sampled and the GG/NEE population sampled, so comparisons between these two projects could be made.
An important part of the pilot PLLCE training in the Northwest was to broaden the scope of training beyond officials who supply good governance to include key members of the public at the commune level who might be in a position to press a demand for good governance. The participants in PLLCE may eventually include many non-officials and thus may contrast with the GG/NEE seminars. However at the time of this evaluation survey, 960 persons had received training but only 72, or 8%, indicated on the attendance sheets that they were not government officials.
Accordingly, at this point, it is reasonable to compare the PLLCE findings with the GG/NEE and Control Group findings, as all these groups consist predominately of urban based government officials. The generally lower scores we found for the commune leaders are probably due to the very different rural background from which the commune leaders come. In addition, the simplifications made in the curriculum delivered to this particular population, may help account for the generally lower scores for commune leaders than for the other groups.
B. Control Group
As noted in the Report, in the sensitive weeks prior to National Elections in July 1998, we were unable to follow the CIHR baseline survey strategy of surveying a gcontrol grouph of randomly selected commune leaders and village notables in provinces where CIHR training was not given.
The aims of this evaluation were to assess the degree of success of the CIHR training in transmitting the knowledge contained in the training curriculum to the trainees. The baseline survey conducted by CIHR aimed to probe general attitudes about the function of the commune leader and was not keyed to specific curriculum knowledge areas. Accordingly it was not feasible to use the CIHR baseline survey as a basis for comparison with our survey findings.
Our Control Group came from among government officials at the same level from which the participants in the Major and Mini Seminars came. The appropriate comparisons are between the Control Group sample and the GG/NEE sample, rather than with the PLLCE sample.
The findings on the Control Group should also be interpreted in light of the way respondents were drawn into the sample. The respondents were located in their offices in provincial towns in the week before elections were to be held. That is to say, they were not out campaigning during working hours, as were many of their colleagues. One might guess that they were not so highly politicized or as zealous in support of a party as their absent colleagues. Their presence in the office in these circumstances probably also indicates that they were conscientious officials.
These respondents volunteered to complete the questionnaire on a completely informal basis, responding to the need of our research team to obtain some findings from the province. The ruling that our team could not officially go into the communes made it necessary to conduct the survey in the provincial town. The willingness of these respondents probably indicates their sense of confidence in their knowledge, although 96% of them indicated that they had not had any courses in human rights or good governance.
We can conclude that the Control Group consists of highly motivated civil servants that probably represent a minority of provincial officials who are not strongly inclined to party politics. These officials had made their own personal efforts to become informed on the issues and subject areas covered in the survey instrument. Their initiative is reflected in their relatively high scores in most knowledge areas.
To facilitate that synoptic comparison of all four groups, a set of graphic presentations of the scores for the four groups is provided below.