Friday, July 22, 2011

Putting the fears puzzle pieces together: social and specific fears in the National Comorbdity Survey

Back in 2008 Bill Tancer wrote a book called Click: What Millions of People Are Doing Online and Why It Matters. Chapter 6 was titled “What are you afraid of and other telling questions.” An excerpt from it was published in USA Today and by ABC News. The first page of the chapter contained a curious list of the top nine groups of fears (allegedly from the National Comorbidity Survey), which are as follows. 

1. Bugs, mice, snakes, and bats
2. Heights
3. Water
4. Public transportation
5. Storms
6. Closed spaces
7. Tunnels and bridges
8. Crowds
9. Speaking in Public

Mr. Tancer says (Ref. #15 in his Notes section) that they came from an article titled “Social Phobia Subtypes in the National Comorbidity Survey,” by Kessler, Stein and Berglund that appeared in the American Journal of Psychiatry in 1998. If you look at the full text for that article you only will find the ninth fear, public speaking, listed or discussed. How come a book about research didn’t check its references? Where did Mr. Tancer really get that list? I don’t know, but suspect it might be from the ECA survey done a decade earlier. Later, I found out that it was.

Mr. Tancer’s list came up again this month in a blog post by Dennis Yeomans, who’d also used it last year. When I asked him in a comment where it came from, he referred to the Kessler, Stein and Berglund paper, which I’d discussed in a blog post back in November 2008.

The National Comorbidity Survey (NCS), is a large mental health survey with data on 8098 people collected from September 1990 to March 1992. That data has been analyzed and discussed in many articles by different teams of psychiatrists. Social fears include topics like public speaking, and talking with others. Other specific fears like heights, flying, small spaces, and water aren’t discussed in the same article that covered social fears. However, a comprehensive survey like the NCS actually does covers both sets of fears (and many other topics), so we can try to put the pieces together ourselves in order to get the big picture.

Table 1 of the Kessler, Stein and Berglund article lists the lifetime prevalence for six social fears. Results are shown above as a bar chart (Click on it for a larger, clearer version).

Another article on “Specific Fears and Phobias: epidemiology and classification,” by Curtis, Magee, Eaton, Wittchen, and Kessler appeared in the British Journal of Psychiatry, also in 1998. You can read the abstract here. In their Table 1 they show the lifetime prevalence for eight specific “unreasonably strong fears.” Their results are added to those on social fears in a second bar chart, shown above. That chart shows the results for 14 fears, just like those from the 1973 Bruskin survey and the 1993 Bruskin-Goldring survey. Note that this list is very different from Mr. Tancer’s, with public speaking at the top rather than the bottom.

It is interesting to compare results from the NCS with those for the same five fears from the 1993 Bruskin-Goldring Survey (in orange), as shown above in a third bar chart. The Bruskin-Goldring results are much higher than those in the NCS. They cannot be rationalized by rescaling (multiplying by the same constant). For public speaking fear their percent is 1.5 times the NCS result, while for height it is twice the NCS. For flying it’s 1.7 times the NCS, for water it’s 3.5 times the NCS, and for being alone it’s 3.2 times the NCS. I think the NCS results probably are closer to the truth.    

It’s best to look it up before you speak up. Sometimes that’s relatively easy, like finding the full text and .pdf for the Stein, Kessler, and Berglund article for free online. The other Curtis et al article was not that easy. To see it I had to go over to the Boise State University library, get a guest pass, find which of their databases had that magazine, and read it.

No comments: