TMIF

The teaching material impact factor (tmif) is a measure of impact for teaching materials (e.g. textbooks or handbooks). It is inspired in citation counting measures as those used to measure research paper or journal impact, but uses as data the references to teaching materials in course descriptions or syllabi instead of citations in research papers.

The ranks based on tmif can also be used for the evaluation of coverage of disciplinary areas in academic libraries.

Basic model
The tmif is a generic model from which concrete measures can be derived. To develop a concrete model for a given discipline $$\mathcal{D}$$ (e.g. Computer Science or Biology), the first step is that of determining a set of knowledge areas $$\{k_1, k_2, \dots k_n\}$$ (following the examples, "Computer Architecture", "Programming", "Information Structure", "Business Organization", etc. or "Animal Biology", "Cell biology", etc.).

Then, the set of course syllabus $$\mathcal{S_D}$$ for discipline D are collected (either a sample or if available, an whole population), and a weighting formula is used to measure the importance of each concrete material (textbook) t in each concrete syllabus s.

The standard option is considering the relative position of the reference in the syllabus as a weight in the interval (0,1]: $$  w(t, s) = min[1, \frac{T_{s} - pos(t) +1}{T_{s}}] $$

where $$T_s$$ is the total number of references in syllabus s, and pos(t) is the position of the textbook been considered.

The tmif then takes the relative importance measures or "votes" and returns an index that can be use for ranking the materials: $$ tmif^{year}(t, k_i)=\sum_{s_j \in \mathcal{S_D}[{k_i, year}]} {w(t, s_j)}\cdot \frac{1}{|\mathcal{S_D}[{k_i, year}]|} $$

Where $$\mathcal{S_D}[{k_i, year}]$$ is the subsets of course syllabi considered for the given year and the given knowledge area (in the domain $$\mathcal{D}$$ considered). Note that the use of a timespan qualifier fits the elaboration of the index yearly, as many higher education institutions renew their course syllabi yearly.

This basic mathematical model can be extended or changed in many different directions, so it is possible to build alternate indexes and empirically test their properties as metrics.

Limitations
The tmif measure can be used to assess the relative impact of teaching materials, but it is only completely meaningful for areas with a common language (as Spanish or English) and reasonably similar curricular structures.

An issue that might impact the measure is the consideration of different editions of the same teaching material. In general, the material must be consider the same if no substantial change in the contents has ocurred from one edition to the next. However, this in some cases has a degree of subjectivity.

Even though the studies carried out so far contain data of traditional, printed handbooks, the same measures could in principle be used for on-line educational materials.

Publications
Sicilia, M.A., Sánchez-Alonso, S. From course bibliography to the evaluation of library collections: the case of Computer Science in Spanish Universities. (submitted)