Posts Tagged ‘Public Library of Science’

Are Facebook Users More Satisfied with Life?

September 15, 2013

This question has been answered in a study published in the Public Library of Science by Ethan Cross of the University of Michigan and Phillipe Verduyn of Leueven University in Belgium. They recruited 82 Facebook users in their late teens or early twenties. Their Facebook activity was monitored for two weeks and each participant had to report five times a day on their state of mind, and their direct social contacts (phone calls and meetings with other people).

The results showed that the more a participant used Facebook in the period between the two questionnaires, the worse she reported feeling the next time she filled in a questionnaire. The participants rated their satisfaction with life at the beginning and again at the end of the study. Participants who used Facebook frequently were more likely to report a decline in satisfaction than those who visited the site infrequently. However, there was a positive association between the amount of direct social contact a volunteer had and how positive she felt. So socialization in the real, as opposed to the virtual or cyber world, did increase positive feelings.

So why was socialization in the cyber world making people feel worse? This question was addressed in another study at Humboldt University and Darmstadt Technical, both of which are located in Germany. They surveyed 584 Facebook users in their twenties. They found that the most common emotion aroused in Facebook is envy. Comparing themselves with peers who have doctored their photographs, amplified, if not lied about, their achievements resulted in envy in the readers.

The question remains whether the same results would be found in older Facebook users. In other words, does age make us wiser?1

1Get a Life! The Economist, April 17, 2013, p.68.

Public to Get Access to U.S. Research

May 1, 2013

This was a title of an article in the Washington Post.1 This news is long overdue. Most scientific professional publications are available through publishers and professional organizations. Usually there are discounts for members of professional organizations, but even we usually pay. I do have access to those published by societies to which I belong. Often, there is an article that I would like to read in a publication to which I don’t have access. Sometimes the fees to access these articles are $30 or higher. It is understandable that publishers, who are in the stated business of making a profit, have such charges. But the charters of most professional societies typically state that one of their objectives is to spread technical knowledge. I hope the irony is obvious here.

Bear in mind that the vast amount of this research is funded by the federal government. So we taxpayers are paying for this research. Then why don’t we have ready access to it? According to the article, agency leaders have been directed to develop rules for releasing federally backed research within a year of publication. Some argue that there should not even be a year’s delay in releasing the information. I agree with these people, but my priority is on the implementation of some policy, and I am against any lengthy debate that would delay implementation.

Aaron Swartz was a genius. He was a brilliant programmer with a list of accomplishments, one of which was the development of Reddit, one of the world’s most widely used social-networking news sites. Two years ago, he was indicted on multiple felony accounts for downloading several million articles from the academic database JSTOR. Although it is not known what his motivation was precisely, one idea is that he intended to upload them onto the Web, so that they could be accessed by anyone. Aaron Swartz was a brilliant and sensitive individual. He was indicted by the federal government and subsequently committed suicide. The March 11, 2013 New Yorker (beginning on page 48) does an admiral job of characterizing this fascinating and interesting individual.

This is more than an issue of fairness. The ready access to this information will benefit both science and the economy. An example cited in the Post article was about a teenage scientist, Jack Andraka, who relied on open access articles to develop a five-minute $3 test for pancreatic cancer.
Fortunately, he was successful, but the charged-for article were an obstacle to his progress.

It should be mentioned that progress has been made in this area. Since 2003 there has been a Public Library of Science (PloS). The healthymemory blog has cited publications from this source and finds it most useful. But this progress has been too slow. This is just another example of how extreme economics has plaques us (See the healthymemory blog post, “Extreme Economics.”)

Similar problems exist regarding the costs of books and higher education, but I’ll stop here before I begin that rant. Enter “higher education” into the search block to read previous rants.

1Vastag, B., & Brown, D. (2013). February 23, A5.

© Douglas Griffith and, 2013. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Douglas Griffith and with appropriate and specific direction to the original content.

Searching for and Evaluating Scholarly Articles

July 18, 2012

 If you are looking for scholarly articles on a topic, Google has a dedicated search engine, You can set up alerts to learn of new articles on topics or authors of interest. Although one can expect and will usually receive, higher quality information from authors extremely knowledgeable in their respective areas of interest, there are certain realities that should be understood. Refereed articles are articles that have been reviewed by several authors knowledgeable in the topic prior to publication. Prior to the internet era this refereeing was needed because paper publication was costly and journals needed to be selective. Typically, there were large delays between the submission of the article, its acceptance, and its eventual publication. There were also journals that did not use referees, that would publish articles for a fee.

With the advent of the internet, the cost of publishing articles and the time to publish articles have been drastically reduced. Yet the archaic artifacts from the print era persist. An author can disseminate an article as soon as she deems it worthy. Why delay the dissemination of information? Some argue that refereeing is still necessary. I’ve participated in this review process both as an author and also as a reviewer. I’ve had articles accepted, and I’ve had articles that I thought worthy that were rejected, but subsequently published by another journal and another review process. As a reviewer I’ve seen articles that I thought worthy of publication be rejected. I’ve also seen articles accepted that, while adequately done, made a questionable contribution to substantive knowledge. Published statistics on the review process are not impressive. Statistics on agreement among reviewers have typically been low. There is also a bar that the editor needs to set to assure that accepted articles have enough allowable pages to accommodate them. I find it odd that academics tend to be impressed by high rejection rates rather than forlorn about research that has gone unpublished. Academics also are keen on refereed journals. Personally, I can quickly ascertain whether an article is worth my time and I don’t need some reviewers editing or censoring the information that is available to me. I think one of the primary reasons academics are keen on refereed journals is that they can use the number of publications in refereed journals in making decision about whom should be awarded tenure. Otherwise, they might actually have to read articles written by tenure candidates.

I see little justification for the traditional institutions for publication. Research can be disseminated quickly via the internet and judgments made regarding the value of the research and on whether it should be ignored or put to good use. The problem is that big moneyed interests are involved. They are the publishers and the professional associations that sell publications. Typically authors and reviewers are not paid. Their efforts are pro bono. The editor might be given an honorarium, but the amounts of small. But the journals and professional books are expensive. And there is no need for them to be. They provide little of added value.

I am able to get online access to journals published by the professional organizations to which I belong for nominal rates. Others are typically quite expensive, as I’m sure some of you can attest. There is a quality online journal that anyone can access for free, PloS 1, Public Library of Science,, this is a referred journal. We should be able to access any research done with taxpayer support online for free. And I, personally, would be willing to forgo the referring requirement.

You should also be wary of biases in different academic disciplines. I’ll provide a couple of examples from my discipline, psychology. When I was a graduate student, there was a large controversy as to whether humans could learn to control their own autonomic processes, heart rate for example. Now it was well known and well documented that Buddhists proficient in meditation could do so. However, this did not constitute the appropriate evidence hard-nosed psychologists required. They wanted to see it done by some student fulfilling a course requirement in a one or two hour psychology experiment.

Shortly after I received my Ph.D and began working at the Army Research Institute for the Behavioral and Social Sciences, I tried to replicate an experiment that I had read in the psychological literature. Although I was able to do so, I was only able to replicate the finding in the group that had General Technical (IQ) scores comparable to college students. The vast majority of reseach in cognitive psychology is based on research done with college students. Although one of the fundamental requirements for generalizing statistical results is that the population to which one is generalizing have been represented in the sample participating in the experiment. I have yet to see a finding in a psychology with the caveat that the results should be restricted to those representated in the statistical sample, college students, for example. I am a working statistician and I am constantly amazed how statistical requirements vary from discipline to discipline when the underlying statistics and their assumptions remain unchanged.

© Douglas Griffith and, 2012. Unauthorized use and/or duplication of this material without express and written permission from this blog’s author and/or owner is strictly prohibited. Excerpts and links may be used, provided that full and clear credit is given to Douglas Griffith and with appropriate and specific direction to the original content.