Mountaineering on a molehill - Chasing the Wind by Felipe B. Miranda

(Conclusion)

(2) Fully adopting the entire data weighting system suggested by Mr. Henares in his critique, practically all the original findings in the Rubicon 2 survey report — a copy of which I had freely furnished Mr. Henares’ group earlier — still remained unaffected, with values changing practically all the time within a point of two points of the original value and at most, in one case, by five points but without changing any rank ordering or subverting any analytical point and therefore validating the original reading in the survey report. By way of illustration, the table below might be considered, with comparable values, one marked ORIG as coming from the original sample and the other marked WTD as the modified values resulting from fully following Mr. Henares’ proposed weighting system:

Table I Awareness and Trust Ratings of Personalities (Row %)

Personalities	Big Trust		Undecided		Small Trust
ÂÂ	ORIG	WTD	ORIG	WTD	ORIG	WTD
Estrada	41	39	24	24	35	37
M-Arroyo	19	19	25	26	56	55
Pimentel	42	41	33	36	24	22
Davide	37	36	38	40	19	18
Belmonte	22	23	41	40	31	33
Aquino	29	30	22	24	49	46
Ramos	27	27	22	22	50	50
Singson	20	18	22	26	57	55

Clearly, there is no significant difference indicated by switching from the original Rubicon report’s findings and using the Henares’ weighting system suggested by way of a critique. This observation holds for all other data in the Rubicon 2 report, whether one speaks of presidential approval/disapproval, agreement/disagreement with presidential resignation and other probes. From this simple demonstration, there can be no reasonable ground for thinking that there had been any manipulation of survey sampling and data generation to deliberately favor any partisan interest, whether pro or contra administration.

There are a few other points raised in the Henares critique like imprudent analyses of too few respondents in a subsample (e.g. Class AB, socioeconomically the best-off respondents, with about 1% — 5 respondents — ) and questionnaire wording.

Perhaps Mr. Henares missed the Rubicon 2 Report’s Table 3 on page 8, where there is delineation of how small this subsample of Class AB is. Given the truly small number of respondents involved here, I do not offer an analysis of this group at all although I still retain the Class AB figures for purposes of indicative analysis (i.e. basically to provide the reader/client possible, tentative hypotheses which might be considered for later, more legitimate probes employing a sufficiently large number of class AB people in a dedicated or a deliberately purposive sample of Class AB members.) In all my presentations of similarly minuscule subsamples, I do not fail to emphasize the simply indicative character of any finding for these respondents.

As regards the possibility of two survey questions relating to rallies being possibly loaded questions (i.e. phrased in such a way as to condition certain responses rather than others), Pulse Asia begs to disagree with the Henares assessment but this matter does not allow for easy resolution and so a full discussion with them and those who might be considered as having some expertise in question design should be around when the discussion takes place. At any rate, the Henares critique does not make this the more critical consideration — as is the case with the sampling procedure used to choose survey respondents — for his press release does acknowledge that the Rubicon 2 questionnaire used "is generally fair."

Critical evaluations of surveys and all their dimensions — their methodologies, sampling and question designs, field interviews, data processing and analysis — are very much in order, especially in these times when so much suspicion attends anything relating to a major political development, the historic impeachment and possible conviction of a Philippine President. Such evaluations, however, must be undertaken with full seriousness and their results subjected first to careful review by competent experts.

The dramatic, presscon approach which works so well in politics and showbiz should be considered only after this serious interactive review has taken place. Otherwise, overly partisan groups might mislead people into thinking that they have uncovered a mountain and induce people to go mountaineering when there is really nothing more than a molehill before them.

Mountaineering, as the critics of Pulse Asia’s Rubicon 2 survey know only too well, is a most dangerous undertaking. Perhaps they should not rush to engage media before they have fully reflected on their critique.