Sunday, May 2, 2010

survey says.....

A few days from now, malalaman na natin kung gaano katotoo ang mga opinion polls na nakita natin during the past few months. If after consistently leading various surveys, it really is Noynoy who will be the next President. O mag-iiba ang resulta at biglang sumigaw si Dick Gordon na tama sya at dapat talagang i-TRO ang mga survey na yan.

Having had a couple of jobs in the market research field during my younger days (ehem… tanda na talaga!) medyo may insight ako sa mga survey-survey na yan. I know that some things are going on in the background na hindi natin basta nakikita. Some things that can really affect the veracity of the reports that the public are being fed.

.

I wouldn’t name names dahil wala pa akong pambayad sa lawyer in case I get sued. But I’d be talking about some recognizable names when it comes to market research in the Philippines. Two companies during that time na swerte naman at pareho kong napag-trabahuhan dati. Based on my experiences with them, I’ll highlight two points at bahala na kayo to judge for yourself kung gaano ba talaga ka-credible ang surveys and opinion polls. Here goes….


Point 1: The hand that rocks the pen


Twas mid '80s when I got hired by Company B as a field researcher. Ang gagawin ko lang, mag-interview ng mga MIS officers ng mga top companies sa Pinas, most of whom are concentrated in the Ayala Area. Pa-survey kasi ng isang international computer giant yon.


Sa isang tulad kong malakas ang loob, walang hiya (notice the space in between ha!) at medyo makapal ang mukha, very ideal yong trabaho. Binigyan ako ng database ng contacts – names, companies and positions – ng aking prospective respondents. All I have to do is to set an appointment with them, show up in time, do the interview and presto, may 150 pesos na ako! Nakaka-2 or 3 interviews ako a day. Malaking pera na yon back in ’85.


The hitch is, hindi mo pwedeng i-submit yong questionnaire na may isang question na blangko (unanswered). It will be nullified at hindi ka babayaran. Ang problema, dahil puro mga executives ang respondent ko, lahat sila limitado ang oras, laging nagmamadali at umaangal sa dami nong tanong. Ewan ko ba naman kung bakit ang daming tanong. Kaya yong ibang interviewee ko, nilalayasan na ako at the 45th question. Ayaw na ring magpa-follow-up interview sa akin. And I’ll be left in dismay.


Eh ano pa ang gagawin ko, di imbentuhin yong sagot sa ibang tanong. Aba, sayang ang effort ko kung hindi ako babayaran. Napudpod ang daliri ko sa telepono kakatawag sa mga supladang sekretarya. Nagbihis ako kuntodo necktie at nag-taxi from Malibay to Ayala. Tapos masasayang lang dahil hindi nasagot yong 5 questions out of 50. Unfair di ba. Kaya ginagawan ko ng paraan. Although I know, a big 'cheat' sign was plastered up my forehead during that time.


So what’s my point?


Surveys are largely dependent on the integrity of the survey itself. Lalo na sa mga taong gumagawa noong actual survey. Kung marami sa researchers ngayon ay tulad kong marunong ding mangdaya, maaring malayo din sa katotohanan ang resulta nong survey. Lalo na kung walang fool-proof mechanism yong survey at madaling manipulahin sa umpisa pa lang.


Point 2: Sino ba ang nagbabayad?


I must have done a good job in Company B kaya nong nag-apply ako sa Company A, I got a recommendation from my previous boss kaya madali akong natanggap. This time, office work na. Hindi kasing-dali nong trabaho ko sa Company B, hindi rin kasing-laki ng bayad pero hindi na 6-months contract ang labanan. This one’s a lot more stable dahil 8-hour job na. And this time, may sarili na akong desk in an airconditioned office.


I was assigned in the consumer survey division. Subscriber namin ang ilang manufacturing company sa Pinas. Companies who wanted to see how their shampoo, bath soap, detergent bar, toothpaste, etc are doing in the market. Kung anong brand, size, packaging at promotional campaign ang bumebenta at hindi. Kaso, I think they are not seeing the real figures they’re supposed to see.


Data Inspector ang tawag sa akin pati sa mga kasamahan ko sa aming maliit na section. Ito yong section na gumagawa ng dirty job dahil kami yong unang binabagsakan ng data after ma-input sa computer ang resulta ng survey ng mga field researchers na nakakalat sa mga department stores, grocery at sari-sari stores sa Greater Metro Area. And what do we do with the data? Hindi lang basta ini-inspect. Ina-adjust pa.


Para kaming mga doctor na nagda-diagnose ng pasyente kung may lagnat o wala. Kung normal yong trend ng data base sa previous survey period, okey lang. Pero kung may lagnat (a spike) yong data na biglang tumaas o kaya nag-chill at biglang bumaba, eentra kami. We will do the calculations para ma-minimize ang gap, ke plus or minus, at mailapit sa dating data point.


Noong ginagawa ko yong trabaho, I was also wondering kung bakit kailangan gawin yon. But my officemates who studied (and graduated BS Stats) explained the principle to me. Hindi pwede ang sharp curves sa graphs. Kailangan laging may statistical regression. Ahhhh… ganon pala yon. Although at the back of my mind the word “cheat” is flashing once again like a bright neon sign.


Eh bakit kailangang gawin? Kasi we have to please the one who pays for the survey. Matutuwa ba sya na biglang nilampaso ang product nyang top seller ng isang bagong produkto? Syempre hindi. So even if the real picture is ugly, we have to give it to the subscriber bit by bit. Hindi pwedeng biglaan. Baka biglang i-pull out yong subscription.


And that’s exactly my second point.


May bahid ng katotohanan ang accusations ng ilan na ang resulta ng survey depends on who pays for it. Sa dati kong trabaho, ang nagbabayad ay yon mismong mga manufacturing companies. And the data we must present them should at the least ‘suit’ their taste.


There you go. Based on experience lang ang kinuwento ko sa inyo. Kaya wag kayong magtaka if I don’t take surveys seriously. Unless I know that the survey is done in all honesty. Tulad nong survey natin dito sa DS. Ito talagang 100% accurate. All the rest, we’ll just have to take it with a grain of salt. Kahit pa rock salt yan. Dahil hindi natin alam kung gaano talaga ka-accurate yong result. And we’re left to believe what we want to.

No comments: