Using a gestalt to measure the quality of in vitro responses.

Authors: Niedz, Randall; Hyndman, Scott; Evens, Terence

Submitted to: Scientia Horticulturae
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/7/2006
Publication Date: 4/23/2007
Citation: Niedz, R.P., Hyndman, S.E., Evens, T.J. 2007. Using a gestalt to measure the quality of in vitro responses. Scientia Horticulturae. 112(3):349-359.

Interpretive Summary: Overall “quality” in plant tissue culture is sometimes difficult to measure. It is customary to measure individual characteristic such as numbers of shoots produced, the size of shoots produced, how many cuttings form roots, the size and number of the roots, etc. However, overall quality is sometimes an unknown function of an unknown number of individual factors. This is the idea of the gestalt where the whole is perceived to be greater than any of its individual parts. We compared four measured characteristics to a gestalt measure where a tissue culture was subjectively scored as “high,” “low,” or “intermediate” quality. The gestalt measure was as statistically robust and accurate as the measured characteristics. We concluded that the gestalt measure may be a useful characteristic in research where quality determinations are difficult to measure but easily identified.

Technical Abstract: Overall “quality” of in vitro responses can sometimes be difficult to assess using measured response variables (e.g., shoot number and height, and callus fresh weight). Gestalt Theory is the idea that the whole is perceived to be greater than the sum of the individual parts. To determine if a gestalt assessment could be used to assess quality of in vitro responses two plant tissue culture systems were examined - Brugmansia x candida shoot multiplication and sweet orange nonembryogenic callus growth. The gestalt assessment of each system was compared to measured responses - shoot number, leaf length and width for Brugmansia x candida, and fresh weight accumulation for citrus. The gestalt analysis modeled as precisely as the measured response variables for both in vitro systems while satisfying the statistical assumptions necessary for a valid analysis. We concluded that the gestalt response is a valid response as it was indistinguishable, in terms of statistical quality, from the measured responses.