*Phase II Survey Instruments
* Group Discussion Specifics
 
You are here: Home > Data > Spatial Data > Spatial Survey Data > Phase II Data Collection > Phase II Survey Instruments > Group Discussion Specifics
Group Discussion Specifics

Team Design and Function – During the course of the fieldwork three teams typically traveled to villages to perform Group Discussion and GPS Accuracy Check tasks while a fourth team remained at Nang Rong headquarters to do Matching.

Group Discussion teams consisted of three members:
  1. The Group Moderator was responsible for introducing the informants to the project, describing the map sets and getting the informants oriented and comfortable with the maps, and for leading / facilitating the group discussion interview.  The moderator also drew plang boundaries, or aided the informants in drawing plang boundaries, and assigned IDs to each plang.  The moderator provided the data recorder with user names, plang IDs, villages names, and map IDs.
  2. The List Manager was responsible for maintaining documents and performing necessary document queries and tallies during the group discussion.  The list manager’s documents included the Household Directory for the target village, copies of the Household List and Dwelling Unit GPS Forms, and a Tally Sheet with the total number of plangs reportedly used by each household in the village.  The Tally Sheet was generated from information found in the Household Survey Form 6 for that village.  The list manager also provided the moderator with ID stickers and verified the accuracy and existence of village user names provided by the informants.
  3. The Data Recorder was responsible for listening, recording, and double-checking all interview information in the Group Discussion Land Use Form and informing the moderator and list manager of missing information.  The data recorder and the list manager were instructed to sit next to each other so as to avoid transcription errors between the Household Directory / Household Listing and the Group Discussion Form.
Orientation Session
  1. Each Group Discussion team, in order to familiarize themselves with a village, reviewed the map tiles and the HH Directory and Household Listing with the field manager prior to going to each village.
  2. After introductions to the project and team members, the moderator described the group discussion task to the headman and his informants and how they would be able to contribute to completing the group discussion tasks.  It was also made clear that the headman was not required to be present for group discussions.
  3. The moderator then read the consent form, and the informants concurred and signed the consent forms, or the moderator signed for the informants if they were unable to sign or if they were wary of signing but verbally consented.  The moderator also signed the consent form as proof that the consent was read to the informants.
  4. The interviewers then completed the Section 1 Cover Sheet of the Group Discussion Land Use Form.
  5. The team then laid out the nine map tiles, explained the dimensions, units, orientation, village center, cadastral lines, etc.
  6. The team moderator asked the informants to stand around the periphery of the maps, close their eyes, and imagine they were flying in an airplane over the village.  The field manager found that this technique significantly helped the informants visualize the landscape from above and think about things geographically.
  7. The informants then opened their eyes and slowly started identifying landmarks and other cultural features, which the team members drew on the maps using China markers and labeled using stickers.
  8. One team member completed labels as quickly as the informants called out the names of roads, temples, ponds, etc.
  9. After landmarks were drawn and identified, each informant was again asked to imagine that they were flying in an airplane over the village, and to fly over and identify their plang from the air.  This was done to see how comfortable they felt with the maps.
  10. All informants were then asked to identify on the maps all major areas of cultivation used by the villagers and to consider who would be a knowledgeable person(s) to invite to discuss the boundaries and users of plangs in those areas in subsequent group discussion sessions.
  11. Team members assigned sequential numbers to the cultivation areas.  Major cultivation area boundaries were drawn on the maps, using identifiable landscape features such as roads, rivers, and canals as reference marks.  The total number of cultivation areas represented the number of ~half-day group discussion sessions that would be required for that village.
  12. Team members then recorded the numbers and the names and ban lek ti’s of possible informants relevant to each area.  Team members inquired with the headman and assistants about the availability of those informants for subsequent group discussion sessions.  The additional informants were contacted later that day or the next day to schedule their participation in a group discussion session.  The intention was to have people in the discussion group more familiar with the areas in question.  These villagers would ideally be selected so that they farmed areas that were spatially well distributed across the cultivation area so as to get better spatial coverage of “local user knowledge”.

Data Collection: Cadastral Mapping and Form Completion
  1. Once the initial orientation session was complete, the spatial team began collecting land use data by asking the headman and the informants to identify on the maps the plangs that they used themselves or that they were readily familiar with, in terms of boundaries and users.  Target plangs were identified and non-target (0-jump and 1-jump) adjacent plangs were identified as needed.  If the informants did not know the username or village name for non-target plangs, the interviewers indicated this with a “Don’t know”.
        Interviewers recorded the user’s full name, however, if the informants did not know the user’s full name, then a common name was acceptable.  Getting the common name should have been rare since the Phase II survey teams were using the household directories to look up villagers’ names.  However, if a person using a non-target adjacent plang was from another village, the common name may have been the only name available and was therefore recorded.

NOTE: Early on in the Spatial Phase 2 Field Work, in order to improve the match rate between the Form 6 data and the Group Discussion data, the interviewers began asking the informants about owner names after inquiring about user names. User and owner names were identified on the maps as described below in the Group Discussion Mapping Specifics section. There were three main reasons for this modification in the methods:
    • During Phase I, there was some confusion among the household interviewers and interviewees about how user/owner related to the definition of “use” in Form 6.  For example, in Form 6, use included growing crops, raising farm animals, or other farm income for the household.  Some interviewers and interviewees equated “other farm income” with renting land to others to grow crops because they were generating household income (via rent) from others growing crops on their land. In this case, the owner was identified as, or identified himself/herself as the user.
    • In Form 6, many of the adjacency names were owner names rather than user names.
    • Many of the group discussion informants had a difficult time separating the user from the owner (problem with user/owner meaning) and/or recalling which name was the owner and which name was the user.
  1. The Moderator’s role was to keep the group discussion moving along, guide the informants over the maps, draw/redraw plang boundaries where necessary, assign IDs to each plang where necessary, and inquire about the users and owners of the plangs.
  2. Once the informants identified an existing plang boundary or helped the moderator draw in a new plang boundary and then identified the user and owner, an existing plang ID or a preprinted sticker ID (beginning with “88---“) was assigned to the plang.
  3. The List Manager provided the Moderator with parcel ID stickers, if necessary.  In addition, the list manager noted the user name, located the user name in the Household Directory, and provided the Recorder with the 9-digit Household ID and the 2000 CEP number for the user.
  4. The List Manager also noted the 3-digit House ID (contained in the 9-digit Household ID).  Using the tally sheet, the List Manager kept a running count of the number of plangs identified for each target village household.  Also recorded on the tally sheet beside the appropriate House ID was the map number and parcel ID for each plang identified.
  5. The Recorder recorded an entry in Section 2 of the GDLUF for each plang/parcel. (see Example 7, Section 2).  Each entry consisted of the Parcel ID, the first and last name of the user, common name of the user, 9-digit Household ID, 2000 CEP number, Map ID, village name, and village ID.  NOTE: The Adjacent Parcel/Plang ID columns in Section 2 – GDLUF were not entered during the group discussion process.  Instead, the matchers completed these entries when they started the matching process for each village.
  6. Sometimes during the course of the group discussion session, the informants were uncertain about plang boundaries and users.  When this occurred either an informant went to the fields and made a sketch map (later transferred to the laminated maps) while the group discussion continued or the interviewer(s) and the informant(s) carried the laminated map(s) to the field for verification.
  7. Sometimes the informants were uncertain about the user or location of a plang, AND they could not visit the fields or households to investigate during the day.  These cases were noted during the group discussion, and the informants took this list with them at the end of the day and did some “homework” in the evening by asking the user or someone who knew about the user’s land.  Any new information obtained was returned to the interviewers the following day.
  8. In many instances, knowledgeable informants were working in the fields harvesting rice.  The interviewers adapted to this situation by going to the informants, either holding the group discussion sessions on the side of the road or under the small shelters that the farmers have constructed in the rice fields.
  9. New households.  If new households were identified during the discussion that were not previously identified or surveyed during Phase I, these new households were assigned an identification number from a sequential list called the “Running List of New House Numbers (House #).”  The new House #s began with a "5" and included a sequential 2-digit code starting with “01”: (“5 _ _”).
  10. After the informants identified and updated on the maps all target and adjacent plang boundaries and corresponding users that they could recall, the moderator drew their attention to areas outside the 9 map tiles.
  11. To begin, the informants were asked to recall any areas outside the 9 map tiles where 5 or more households used land for cultivation.  If any such areas existed, they were asked to identify the household and the number of plangs used by each household in addition to the village that these areas were located near and the sub-district.  The N-map was used as a general reference to help the informants determine which village the plangs were near.  The relevant information was recorded in Section 3 of the GDLUF.
  12. Finally, the informants were asked to recall any Other areas (areas where 4 or fewer target village households used land, including single plangs) outside the 9 tiles where households used land. The relevant information was recorded in Section 3 of the GDLUF.
  13. This concluded the first group discussion session. Subsequent group discussions followed the same format as steps 1-14.  The headman was not required to be present at any of the group discussion sessions. For subsequent group discussion sessions, a new major area of cultivation was focused upon, and the informants most familiar with that area were called upon to assist in the group discussion.
Group Discussion and Form 6 Comparison of Number of Plangs Used
  1. Following each group discussion session, the List Manager compared the entries in the GDLUF (completed by the Recorder) to the tally sheet to check his/her running tally.
  2. After all group discussion sessions were completed and the tally sheet was checked and finalized against the GDLUF, it indicated that either less, the same number, or more plangs were identified per household during the group discussion when compared to the numbers reported in Household Questionnaire Form 6. (Recall that Form 6 number of plangs used for individuals were transferred to the Household Directory and totaled per household in the Tally Sheet).
  3. For each House ID on the tally sheet, if more plangs were identified in the group discussion than reported in Form 6, then that House ID was circled.  If the same number of plangs were identified in the group discussion as reported in Form 6, the House ID was also circled.
  4. If a smaller number of plangs was identified in the group discussion compared to the number reported in Form 6, the interview team assembled a final group discussion session to address these discrepancies. This final group discussion session was made up of as many knowledgeable informants as the interviewers could find.  Since the team had already written the names of the heads of households beside each House ID, they simply stepped through the list on the tally sheet.  For each non-circled House ID, they inquired as to whether there were any additional plangs used by the head of household or members of the household.
NOTE: The total number of plangs reported as being used by this household from Form 6 was not disclosed to the group discussion assembly.
  1. Typically, four outcomes resulted from this closing group discussion session:
    • The informants had overlooked these plangs during a previous group discussion, in which case the updated boundaries were drawn, IDs were assigned, and entries were recorded in the Land Use Form and tally sheet.
    • The plangs in question were overlooked and located outside the nine tiles, in which case the relevant information was recorded in Section 3 of the GDLUF.
    • The informants did not have any additional information regarding a household’s plangs.
    • The possibility that plangs may have changed users during the last cultivation year.  For example, the group discussion informants were reporting that one person used a plang that another person reported that he/she used in Form 6.  If so, this information was noted on the tally sheet and GDLUF.

Group Discussion Mapping Specifics
        Materials used for mapping included china markers, rulers, paint thinner and tissue for removing "permanent" ink, label stickers, and clear tape.  All teams used the following mapping and map annotation conventions when collecting data:
  1. Black/Brown Lines = Paved roads, Dirt roads, paths, and cement roads
  2. Blue Solid Lines = Permanent streams, canals, creeks, ponds, swamps
  3. Blue Dashed Lines = Intermittent streams, canals, creeks, ponds (rainy season)
  4. Red Lines = Updated, current plang boundaries
  5. Red Circles around mapped plang IDs = indicated plangs used by villagers from target village
  6. Green Circles around mapped plang IDs = indicated plangs used by villagers from all non-target village
  7. White, Rectangular Plang ID stickers = indicates the new 5-digit plang ID assigned to an updated plang. For each target village the ID stickers were sequential beginning with 88001.  This assured unique plang IDs for each village group discussion
  8. Red Hatch mark on Plang ID stickers = indicated plang used by villagers from target village
  9. Green Hatch mark on Plang ID stickers = indicated plang used by villagers from all non-target villages
  10. Yellow Round sticker inside Plang = indicated that plang was matched to a plang in the Form 6 data
  11. Yellow Highlighter shade on Plang ID stickers = indicated that plang was matched to a plang in the Form 6 data
  12. Name Written in Red inside Plang or on the sticker = user name who lives within target village
  13. Name Written in Red inside Plang or on the sticker and in parentheses = owner name who lives within target village
  14. Name Written in Green inside Plang or on the sticker = user name who lives in non-target village
  15. Name Written in Green inside Plang or on the sticker and in parentheses = owner name who lives in non-target village
  16. All stickers were covered with scotch tape to prevent peeling
  17. The original map – tile IDs on the maps consisted of the 4-digit 1994 village ID and a tile number (with the exception of the 16, year 2000 splits, which had 6-digit IDs).  The 4 digit 1994 ID’s were replaced with the 6 digit 2000 ID’s and tile number using preprinted stickers
  18. All correct pre-existing cadastral lines and updated cadastral lines were traced in red permanent ink only after the final boundary was determined and agreed upon by all the informants present
  19. Parcels that were completely contained within one map were to be spatially complete when drawn, that is, drawn with no gaps in their boundaries or edges.  They formed closed polygons
  20. If a single parcel was split by two map tiles, the parcel boundary was drawn to the edge of the map tile and the resulting partial parcels were assigned the same “88---“ ID, indicating that those two pieces make up a single parcel
  21. Parcels located at the corners or edges of the 9-tile map area (36 square kilometers) that were not spatially complete (i.e. the polygon boundary extended beyond the map edge) were included in the data collection if they were target plangs or were adjacent (0-jump and 1-jump) to a target plang
  22. Some small plangs were smaller than the ID stickers themselves.  In these cases, the sticker was not placed over the small plang, rather the sticker was placed somewhere near the plang and an arrow was drawn from the ID sticker to the corresponding plang
  23. Existing parcel boundaries and parcel IDs: For existing cadastralized areas, informants verified the accuracy of the cadastral boundaries.  This involved making no changes, or marking out existing cadastral lines, or splitting an existing cadastral parcel into multiple parcels, if necessary
  24. New parcel boudaries and IDs: For uncadastralized areas, informants identified the parcel boundaries, which were then drawn on the map using red markers. 
The following caveats apply to 23 and 24:

    • If many individual parcels were combined into a single larger plang, based upon a common user, the new boundary (all boundaries) was drawn using red markers.  The old parcel IDs were marked through with a black marker and a single new ID sticker (with ID beginning 88---) was assigned to the aggregated parcel.
    • If one parcel was split into two or more parcels, the new resulting parcels were assigned new sticker IDs beginning with "88---".
    • Immediately adjacent plangs with the same user were defined as one plang with a single ID.
    • Some existing cadastral parcels were labeled with a "0" or "-9" because the original cadastral maps obtained from Thailand did not include IDs for these parcels.  If these parcels were included as part of the updating process, the "0" or "-9" ID was marked out and replaced, if necessary, with a new parcel ID.
    • There were cases where one large plang was split by a road or canal.  The interviewers left this plang as a single large plang, but gathered user information and temporary boundaries for the two pieces on either side of the road or canal.  Then they checked the information supplied in Form 6 to see if the HH interviewer had recorded the plang as one large plang or not.  If the HH interviewer recorded two separate plangs, then the interviewers made the permanent split on the map.  Some Form 6’s did not split plangs that straddled a road or canal and some Form 6’s recorded them as separate plang.
    • Multiple households using one parcel.  Some plang were utilized by more than one individual/household. Each user of the parcel was recorded in a separate record (row) on the GDLUF, therefore, the same Parcel IDs appeared in more than one record on the GDLUF, but each record contained a different House ID.
    • Parcel IDs were assigned to the public land areas only if the informants could draw in the approximate parcel boundaries within the public land area.


  Last Modified: 04/22/2004 UNC Carolina Population Center