A dream goal for any brand team sponsoring a poll or survey for publication is to have the findings covered by the New York Times so, to help brands achieve this goal, we’ve been taking a look at the kind of polls/surveys that the outlet covers.
During the month of October 2014, the New York Times ran 116* stories that referenced a poll or survey. Here's what they were about; who sponsored them; and how they were conducted:
Stories most commonly supported by data derived from a poll or survey were focused on health and wellness and, in particular, health insurance (N=9), psychology (7), drug use (3), and Ebola (3). The economy - especially economic outlook (14) - was the second most reported topic supported by poll/survey data.
The most typical type of sponsor of a poll or survey covered by the New York Times is an NGO with the Pew Research Center (a subsidiary of the Pew Charitable Trusts) standing out as the most prolific sponsor of polls/surveys reported by the outlet.
Other frequent sponsors include universities and media organizations. There were also a handful of government and private business sponsors with the latter mainly consisting of banks and professional service firms.
Some surveys were referenced in more than one article so were included in the counts/percentages more than once. Thomson Reuters Extel Survey (pan-European research among buy-side, sell-side and corporate communities), for example, was referenced seven times; Hong Kong University’s barometer on democracy in the region was referenced four times; and AP’s preseason poll on college football was referenced three times.
The New York Times reported on polls and surveys adopting a broad range of designs and methodologies. Most of the surveys and polls it reported on were longitudinal in nature i.e. updated at least once. Of the 99 surveys/polls where the timeline was clearly stated, almost three-quarters were an update to (N=68) or the start of (2) a long-term trend.
Of the 48 surveys/polls where the methodology was clear, more than half (26) were conducted online; a third (16) were conducted by telephone (often with the deliberate inclusion of cellphone users); a fifth (8) were conducted in person; and three were conducted by post.
Of the 63 surveys/polls where the sample size was clearly identifiable, half (33) had a sample size of around 1,000 or under including a fifth (12) with a sample size of around 100 or under. The surveys with smaller samples were conducted among elite/opinion leader type audiences such as senior executives, economists, or journalists.
Of the 107 surveys or polls where the geography was clearly identifiable, most were conducted across all (61%) or a region of (10%) the United States. Just over one in ten (12%) had a global focus.
* Analysis excludes surveys and polls pertaining to government/candidate approval and voting intention; mentions in listicles e.g. “Today in Small Business” blog; surveys of anything not human (e.g. buildings); and clinical studies.
Cases were excluded from counts or percentages if more than one type of sponsor or methodology was cited within an article.