Sampling - April 28, 2009 Sampling (Continued) •...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: April 28, 2009 Sampling (Continued) • Representative Sampling (Probability Sampling): - Repr e s e n t a tiv e b e c a u s e e v e r y o n e in t h e p o p ul a tio n h a s a n e q u al c h a n c e of b ei n g inclu d e d in t h e s a m pl e - How r e pr e s e n t a tiv e is it? o Sampling Error: sample data will be slightly different from population because of chance alone; estimate this statistically (margin of error) Smaller margin of error if: • More homogenous population (can’t control) • Larger sample size o Systematic Error (sampling bias): systematically over or underrepresent certain segments of population (going to a DMV to see how people are going to vote – what about non-drivers?) Caused by: • Improper weighting of certain groups; very low response rate (research has an accommodation bias – people who are nice and willing to participate) • Using non-representative sampling methods • Representative Sampling Techniques (Random): Si m pl e Ra n d o m S a m plin g: s el e c t el e m e n t s r a n d o mly fro m p o p ul a tio n o List e d p o p ul a tio n s: u s e r a n d o m n u m b e r s t a bl e o Pho n e s : r a n d o m-digit di alin g Sy s t e m a tic S a m plin g: fro moa t h e p o p ul a tio n, s el e c t e v e r y n t h l ist f el e m e n t o Mus t h a vreando m start e t hro u g h ; cycl e ntire list o Si milar r e s ult s a s si m pl e r a n d o m s a m plin g o Watc h o u t for p o t e n ti al “ p e rio dicity” (listin g b o y, girl, b o y, girl) S tr a tifie d S a m plin g: divid e t h e p o p ul a tio n into s u b s e t s (“s tr a t a ”), a n d t h e n s el e c t r a n d o mly fro m e a c h o Us u ally s tr a tify for d e m o g r a p hic v ari a bl e s (s e x, r a c e, p olitic al p a r t y) o N e e d p rior kn owl e d g e of p o p ul a tio n pr o p or tio n s o Incr e a s e s r e pr e s e n t a tiv e n e s s of s u b gr o u p s (r e d u c e s s a m plin g e rror for t h e s tr a tifie d v ari a bl e) o Mor e c o s tly a n d ti m e c o n s u mi n g Multis t a g e Clus t e r S a m plin g: firs t r a n d o mlyroups “clu s t e r s ”), g s a m( e pl t h e n r a n d o mly s a m pl e el e m e n t s wit hin e a c h clu s t e r o Ex a m pl e: S a m plin g “S t u d e n t Clu b Me m b e r s ” 1st Stage: randomly sample universities - - - 2nd Stage: randomly sample clubs from those universities in the sample 3rd Stage: randomly sample members from the clubs in the sample o Useful for populations not listed as individuals o Reduces costs o Sampling error at each stage Multis t a g e & Str a tifie d S a m plin g Co m bin e d: o Ex a m pl e: S a m plin g “Ch urc h Me m b e r s ” 1st Stage: sample churches, but stratify for religion/denomination (divide into Catholic, Protestant, Jewish, etc) 2nd Stage: sample members • Non-Representative Sampling Techniques (Can’t Generalize!!): - Conv e ni e n c e S a m pl e: s el e c t individ u als t h a t a r e a v aila bl e/ h a n d y - Pur p o siv e S a m pl e: s el e c t c e r t ain individ u als for a s p e ci al r e a s o n (t h eir c h a r a c t e ris tic s, e t c.) - Volun t e e r S a m pl e: p e o pl e s el e c t t h e m s elv e s b e c a u s e t h e y w a n t t o b e i nclu d e d - Q uo t a S a m pl e: s el e c t individ u als t o m a t c h t h e d e m o g r a p hic pr o p or tio n in t h e p o p ul a tio n (filling q u o t a s ) - N e tw ork/S n ow b all S a m pl e: s el e c t individ u als w h o c o n t a c t o t h e r si mil ar i ndivid u als a n d s o o n a n d s o fort h • Use of Time in Studies: - Cros s-S e c tio n al S t u di e s:s a m pl e a t e p oin t in ti m e one on - Lon git u din al St u di e s: m o r e t h a n o n e p oin t in ti m e m e a s ur e d o Pa n el – s a m e p e o pl e e a c h ti m e o Tren d – differ e n t r a n d o m s a m pl e s e a c h ti m e o Coh ort – differ e n t s a m pl e s, b u t of s a m e “co h or t” (no t s a m e g e n e r al p o p ul a tio n, b u t s o m e t hin g t h a t links p e o pl e b a s e d o n s o m e e x p e ri e n c e t h e y’v e s h a r e d t o g e t h e r, u s u ally h a vin g t o d o wit h ti m e – for e x a m pl e, y o ur g r a d u a tin g cl a s s) ...
View Full Document

Ask a homework question - tutors are online