What can be done to bring dead capital to life? There's a list of survey design recommendations (that apply generally to tests and assessments as well) on the last three pages of the issue of Rasch Measurement Transactions posted at http://www.rasch.org/rmt/rmt203.pdf .
As to the nuts and bolts, specialized software for data analysis is available from a number of sources. Ministeps is available for free at http://www.winsteps.com/ministep.htm but works only for very small data sets (25 items and 75 cases). For larger data sets, you'll need Winsteps ( http://www.winsteps.com/winsteps.htm ), RUMM ( http://www.rummlab.com.au/ ), ConQuest (available at www.assess.com), or one of the several other available programs (for a list with web sites, see http://www.rasch.org/software.htm ). If you're calibrating judges assessing employees' performances you'll need either ConQuest or Facets (http://www.winsteps.com/facets.htm). In a multifaceted measurement situation (people rated by judges relative to criteria using a rating scale), you'll need to pay close attention to your assessment design to make sure all the judges are linked (see the Facets manual for more info).
Winsteps includes worked examples from the best-selling introductory book on practical approaches to realizing meaningful measurement and fairly calibrated instrumentation (Bond & Fox, 2007).
Information on training workshops is available on the respective web sites, and at www.rasch.org/confer.htm.. I am available to answer questions and for larger consulting projects, by request.
Obtaining really meaningful, fair, and objective measurement takes considerable expertise. Local instrument calilbrations are not a huge problem, but, just as no single individual, firm, or industry could have created the existing metric system alone, we must all act together to create a system of universally uniform metrics for human, social, and natural capital.
Given the essential role precision measurement plays in reducing transaction costs and reducing market frictions, and that NIST documents returns on investment of 40% to over 400% for various improvements to the existing metric system, I'm betting that health care reform, improved educational outcomes, and effective ways of addressing environmental issues all hinge on creating a common language of meaningful measures we all think in together.
I've been writing to various Senate committees, etc., but my lone single voice is drowned out in the general cacophony.
I am open to suggestions on how to proceed from here.
In principle, three areas of activity need to be addressed simultaneously.
First, the existing metrology infrastructure must be expanded to include measures of human, social, and natural capital. This infrastructure must require:
that all instruments measuring the same thing do so in the same universal uniform metric that is additive, divisible, and mobile,
open architectures that allow and encourage metrological quality improvements,
that the needs of as many stakeholders as possible be met in the common language of a single standard,
the same kinds of checks on the stability of the metrics as are in use in the Systeme Internationale, and
that the measures can be aggregated for use in managing each significant economic level and sector.
The second major area of activity’s primary focus should be on identifying what forms of human, social, and natural capital need to be measured and managed. An incomplete list of possible forms of capital to be measured includes:
Health and functionality
Literacy and numeracy
Motivations and attitudes
Abilities and skills
Mutual trust, good will, loyalty
Quality of life
Air and water purification
Gene pool diversity
Watershed erosion protection
Third, this far-from-comprehensive list will be then complemented by questions as to what experiments have already tested the key forms of capital for additivity, divisibility, and mobility.
Reference cited Bond, T., & Fox, C. (2007). Applying the Rasch model: Fundamental measurement in the human sciences, 2d edition. Mahwah, New Jersey: Lawrence Erlbaum Associates.
The third major area of activity will be the design of tests,
surveys, and assessments to measure what needs to be managed. It will
be vital that instruments be designed so that data will meet the
mathematically rigorous requirements of additivity, divisibility, and
mobility in publicly accessible and reproducible ways. One set of
approaches to this work (described in more detail elsewhere (Fisher, 2005)) that has proven to be of marked utility
comes under the heading of Rasch models (Rasch, 1960, 1961, 1977;
Andrich, 1988; Wright, 1977, 1999; Bond & Fox, 2007;
http://www.rasch.org). These models are implemented via fairly widely
accessible software, are employed in a wide variety of published
research studies, and are made internationally available by
consultants, conferences, and training seminars.
Copyright 1995-2014 Living Capital Metrics Last updated May 9, 2014.