Weekend reads: Sexism from a Nobel laureate; publisher deception; irreproducibility’s price tag

booksThe week at Retraction Watch featured the story behind a Nature retraction, and the retraction of a paper by a pioneer in the field of exosome research. Here’s what was happening elsewhere:

7 thoughts on “Weekend reads: Sexism from a Nobel laureate; publisher deception; irreproducibility’s price tag”

  1. Comment on the Pomeroy’s blog – anyone who has put the time into reviewing the nutrition science literature knows the limitations of short term dietary recall data. The problem is not the quality of the recall but the duration. The “true” average consumption of calories or nutrients may require weeks or months of recall data. Basiotis et al (1987) (http://jn.nutrition.org/content/117/9/1638.full.pdf – free access from publisher) suggests the minimum number of days needed to arrive at an acceptable average for calories (3 days), calcium (7 to 10 days), vitamin A (39 to 44 days) and other nutrients. Their references include a paper from the 1950’s so this is not a recent discovery. According to Google Scholar, Basiotis et al is cited in 461 articles.

    The main use for the NHAMES dietary recall data is to get a sense of the food preferences of the US national population. It is used by US nutritional programs such as WIC, the school lunch program and SNAP. In working with the data, I learned that the typical kid, regardless of race or ethnicity, really likes french fires. Research has long argued that more days of recall are needed to get good estimates of typical food consumption. The barrier to get more days is the cost. The time needed to conduct the dietary recall survey can be 30 to 60 minutes per person per day. To maintain the quality of the survey, it is conducted by a registered dietician.

    The Basiotis et al paper and similar papers are not discussed in the paper (Archer et al (2015)) mentioned in the blog. I am really surprised that the Archer et al article missed these papers. By focusing on only one paper, Pomeroy has contributed to the problem.

  2. I think Pomeroy’s comments are geared more at associations between diet and disease, not just indices of caloric intake or estimating micronutrient status.

    We should have some difficult conversations with journals and their editors. Part of their overhead costs involve editors and various items related to referring articles to experts and reviewers for input and quality control. They are paid large sums of money to cover those costs and their profits seem to be pretty high especially at the big name journals. Maybe it’s time to start grading the journals on how well they’re doing their job and renegotiating their fees. If they want to balk at that then how about all university ORI offices clear each research effort for integrity, validity type issues and then the university sends the manuscript off to another university for review and the reviewing university can serve as reviewers and if they endorse it they can simply put it on their website and charge a small fee to download it and let it grow from there. If the university athletic conferences can create their own networks (like the SEC Network for example) then I’m sure universities can get together and create their own publishing networks.

  3. Tim Hunt:
    Let’s use scientific approach instead: ask Hunt’s female employees for opinion, not his wife or heads of ERC, EMBO, Sainsbury Lab and other senior peers. Let us simply collect data evidence from the most appropriate source: those who Hunt was actually talking about.
    I haven’t seen anyone take this outlandish option so far: ask all the “girls” (who cry when criticized or try to seduce poor Tim Hunt) about their work experience.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.