obviously the first one is more reliable in a statistical accuracy point of view. The larger the sample size the more accurate the information is. From a sample size of 223,518 to 73,940. But these facebook polls are worthless really.
Uhmmm… it’s Sunday and I refused to go to church → hence boredom. Now that we don’ got that outta thaway…
That’s not exactly accurate. A larger sample size is always better but even more importantly is HOW the sample is collected. What measures did you apply to counter measurement, experimenter (survey) bias ? (- moderator bias, biased questions, biased answers, biased samples, biased reporting - all these refer to undue influence on the audience, questions and answers)
The non-anal comment :D:D:D is that we don’t know who visits the site. In other words, your last statement is most accurate. Facebook surveys have very little external validity.
Honestly, kwa hio first poll i would have called them out for using photos that portray two very different things. One candidate is portrayed as easy going and fun loving while the other is seen to miserable and defensive. There are many ways to rig a poll and subliminal messaging works quite well for people who are on the fence. Would the result be the same if they used this photo of Uhuru?
[ATTACH=full]93349[/ATTACH]
And before the Uthamaki group of handles crucify me and tell me that RWNEBP, I’d like to let them know that I already know that