H.A. Goodman deconstructs bogus poll results.

[Admiral Loinpresser]

As for how these polls are conducted, they'd be very relevant in the 1970's. First, the majority of these polls call primarily landline telephones. Sorry, but the latest poll with Clinton 30 points ahead of Sanders states the respondents were "1,000 Adults, including 350 respondents with a cell phone only." Also, the Iowa poll with Clinton up by 41 points asked 300 (out of only 400 people) respondents via landline telephone.

The biggest issue here is that almost half of Americans don't use a landline telephone; especially Americans under 45 years old. A Time article titled Landline Phones Are Getting Closer to Extinction highlights why landline surveys could be skewed towards Clinton's favor:

41% of U.S. households were wireless-only by of the end of 2013, according to new data from the National Center for Health Statistics.

Young adults are unsurprisingly the cohort the most likely to live in wireless-only homes, with 66% of people between 25 and 29 using cellphones exclusively. Americans between 30 and 34 were the next largest group of cord-cutters, with 60% of them living in wireless-only homes. 53% of people between 18 and 24 are now cellphone-only, while 48% of people aged 35 to 44 and 31% of people aged 45 to 64 have made the jump.

Just 14% of adults over 65 have dumped their landlines, though. Overall, more than half of wireless-only adults are now 35 or older, up from 47.6% in the second half of 2010.

Since the demographic groups most likely to use cell phones are likely to vote for Bernie Sanders, it's logical that polls show Clinton ahead when primarily landlines are surveyed.

http://www.huffingtonpost.com/h-a-goodman/landline-telephones-moral-tone_b_8470102.html