The
article I analyzed was Gallup's Congressional
Approval Sinks to Record Low published on November 12th, 2013. The
article highlights how Congress previously held a record low approval rating of
10% in 2012, only to be surpassed this November, falling to 9%. The results
were collected from a survey Gallup has conducted every year since 1974. On
average, Gallup calculates Congress has an approval rating of 33% since it
started collecting data.
For this
survey, Gallup used phone interviews conducted both in English and Spanish, for
landlines and mobile phones. Surveys were done each month in 2013 for 3 to 6
days, collecting between 100 and 101 responses. For 2013, individuals were
selected using a random digit dial method. Random digit dialing generates telephone
numbers arbitrarily for selected areas and has the advantage of including
unlisted numbers that would otherwise be overlooked. Disadvantages of this
method occur when responses to the survey differ from those who were contacted
but unable to respond (De Michele). An example of this could be a night worker
who doesn't answer the phone during normal business hours. There are many ways
to mitigate the non-response rate including multiple phone calls, monetary
incentives, advance letters and varying survey call times (Sangster). If
necessary, landline numbers with multiple potential adult respondents were filtered
by selecting the individual with the most recent birthday.
Gallup then
attempts to correct the sample data by weighing for unequal selection
probability, non-response, and double coverage of landline and mobile phone
users. In addition, these results are then weighted according to national
demographics of gender, age, race, ethnicity, education, religion, population
density, and phone status (mobile only/landline only/both, mobile primarily).
These weighted figures are derived from: March 2012 Current Population Survey, July – December 2011 National Health Interview Survey, and
the 2010 Census.
All data
gathered since 1974 is included in the article. It is unclear if Gallup has
always used the same method of data collection and weighing of the samples. Frequency
of sample collection has varied widely. There are several years (1978 – 1988) in
which Gallup only surveyed the population once; therefore, the average 33%
approval rating is suspect. It also points to the likelihood that Gallup’s
methods, as we would hope, have changed over time. This particular sample is
calculated with 95% confidence and a margin of error ± 4% points.
In my
opinion, most of the movement in this poll can be attributed to high profile
political events. Take for instance the highest congressional approval rating
of 56% in 2001. This rating is due mostly to approval scores of 84 in October,
73 in November, and 72 in December, attributed most likely to American
nationalism following 9/11. In comparison, the political climate of 2013 has
been highly divisive; no survey sample has climbed above an approval of 20.
Issues including the October 2013 government shutdown and problematic roll-out
of federal healthcare reform have clearly led the numbers to new lows.
Article:
References:
No comments:
Post a Comment