Data Collection Consistency

To calibrate the consistency with which university groups used the Instrument to collect data, each university independently reviewed and collected data on the recovery plan for one species, Kokia drynarioides. (Kokia drynarioides is an endangered tree from Hawaii). Ideally, each university should have recorded the same response to each question in the Instrument since they were all reviewing the same plan. Discrepancies among responses would indicate measurement or recording error. Discrepancies could also indicate that universities were interpreting a question differently.

For a more detailed analysis and discussion of these data, see:
Hoekstra, J. M., Clark, J. A., Fagan, W. F. and Boersma, P. D. 2002. A comprehensive review of Endangered Species Act Recovery plans. Ecological Applications 12:530-540.

Important: Any assessments of data-collection consistency inferred from these data must be interpretted cautiously. Taken in aggregate, we believe that the data provide a general measure of consistency among university groups. However, evaluation of consistency regarding specific questions cannot be considered reliable because the sample size for such assessment is one. Statistically reliable assessments of consistency for individual questions in the Instrument would require calibration data from multiple recovery plans.

Use the links below to download data for the Kokia drynarioides recovery plan. Only 18 of the 19 participating universities submitted data for this plan. Representatives of the U.S. Fish and Wildlife Service also collected data for Kokia. These data are included in the files below, but were not included in summaries or analyses of data collection consistency. Due to the large quantity of data, data were split into smaller files by Form from the Instrument. Rows contain data collected by a particular university. Columns contain data from a particular cell in the Instrument referenced by a unique alphanumeric code that corresponds to the row and column identifiers. Data response codes are desribed in the Key to the Instrument. Raw data files are comma-delimited text (.CSV format). The summary file is a Microsoft Excel workbook of spreadsheets (.XLS format).

Raw Data

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

Summary Data

Kokia summaries containing summaries of the Kokia data for each question in the Instrument. The file contains 13 worksheets, one for each Form in the Instrument. Each worksheet contains the following information about each question:
QID = Unique alphanumeric identifier for the question (row number + column letters).

RowID = Row number of the question from Instrument.

ColID = Column identifier of the question from Instrument.

FormID = Form on which the question occurred in Instrument.

Datatype = Categorization of the question according to the type of data coded:
C = Count data (e.g. how many of something)
F = Unordered factor data (e.g. red, blue, yellow)
O = Ordered factor data (e.g. small, medium, large)
Y = Yes/No responses
T = Text responses

Resp.cats = Categorization of question according to the number of possible response codes:
5 (for yes/no questions plus special response codes: 1, 0, -1, -2, -3)
6
7 or 8
9
10+ (e.g for habitat types, or day of the month)

Resp.tables = Tabulations of responses to each question.
Numbers in the top row list the response codes that were recorded for each question.
Numbers in the bottom row give the number of times the response code above was recorded.
These values should sum to 18, the total number of universities that submitted Kokia data.

These summaries were compiled by Jon Hoekstra. If you have any questions or comments regarding these data or summaries, please contact recovery@u.washington.edu.

Back to Top

Go back to the Data Collection page

Go back to the main Recovery page

Please send comments or questions about this site to recovery@u.washington.edu.
Last updated: 9 August, 2002.