Table of Contents Coding
Tables of Contents (TOC) for the various series were coded similarly to the questions in the main body of the texts.
This coding provides some small data sets to begin exploration of multivariate methods from Ecology as they might be applied to the textbook data.
They also provide the opportunity for a consistency check between what can be gleaned from a Table of Contents as compared to the exhaustive scoring of the problems in the body of the text itself.
|
|
|
TOC Scans with codes
Searchable scans of the paper copies of the coded Tables of Contents. These were given to me ( Dick Furnas) in a stack and were scanned from top to bottom of the stack. File names include the automatic time stamp of when the scan was performed, as well as annotations from the first page of the TOC identifying the series and sometimes who did the coding when it was marked on the paper.
Batch 1
Batch 2
TOC Consolidation
Data from the paper scans were entered into a single spreadsheet. Each row in the spreadsheet either provides context for the data or is an actual data row. Context is provided in a hierarchy which might be thought of as an outline form:
Series Book |
|
|
|
|
|
|
|
Chapter |
|
|
|
Description |
(Detail) |
|
|
Section |
|
|
Description |
(Detail) |
|
|
|
Code |
Importance |
|
|
|
|
|
Code |
Importance |
|
|
To use the data, the context can provide the basis for groupings of codes to correspond with the notion of "Sites" in the context of Ecological data. The most obvious analogies would be to associate sites at the series or chapter levels. Section is probably too detailed for the TOC analysis, since some series do not have a notion of section and those that do, usually had only a single code applying to the section.
Codes recorded as written on the paper. By convention, where codes were circled, importance was taken to be
2 , otherwise
1 . In the absence of circled codes, importance was taken to be
2 . Where a code had a question mark next to it, its importance was taken to be
1 . The code
999 was used when a code was missing or unintelligible. Such entries were given importance
0.
Coding Conventions for data used in Ecological Software
For each series we have coded multiple books:
- CPMP Course 1, Course 2, Course 3
- IMP Year 1, Year 2, Year 3
- CME Algebra 1 & Algebra 2
- Discovering Mathematics Algebra & Advanced Algebra
- UCSMP Algebra & Advanced Algebra
- Glencoe Algebra 1 & Algebra 2
Short (up to 8 character) Codes for books/coders.
Codes are in the format:
- BOOKXSC * i.e. [book (4characters)][chapter X (1 character)][section S (1 character)][coder C (1 character)]
- CPMP: Core Plus Mathematics Program
- IMP:
- CME:
- Discovering Mathematics:
- UCSMP:
- Glencoe:
Comments
- Labels for the Ecological software are ideally at most 8 characters long. We should arrive at a consistent coding convention to label whatever we explore as "Sites" so the provenance of the data is clear. Items which seem relevant include: Series-Book;Chapter;Section;Coder which suggests two characters for each. -- DickFurnas - 05 Jan 2013
--
DickFurnas - 2013-01-05