For the article below:

Is the Kross research an experiment? Why or why not and what type is it? Why is the Kramer research described as an experimental design? Were the 3 threats to internal validity avoided by the Kramer research? Why or why not? Is there manipulation and random assignment? What is the independent variable, dependent variable, mediating variable and moderating variable?

Article: How Facebook use affects wellbeing

"Here we are at the fabulous Hibiscus Inn, sipping daiquiris and overlooking the blue, blue Caribbean Sea." "Our daughter, Angela, just graduated at the top of her MBA class in June. Angie will be starting her new job next week, and we couldn't be prouder of her achievements." Ever notice how your Facebook friends are happier, luckier, and more successful than you are? Researchers have found a correlation between Facebook use and negative emotions. In particular, the more time people spend perusing their friends' Facebook pages, the more they experience feelings of jealousy, loneliness, and low self-esteem. In 2013, researcher Ethan Kross and his colleagues published the results of their research into this phenomenon. In a longitudinal study, Kross et al followed 80 Facebook users for two weeks, monitoring their usage of the social networking site. Participants were sent text messages five times a day, in which they were asked to describe their feelings of wellbeing, worry, and loneliness on a survey questionnaire.

They also reported how much they had used Facebook since their last response to the questionnaire as well as the frequency of their direct, face-toface and telephone interactions with others. In another measure, participants rated their life satisfaction at the start and at the end of the 14-day period. What emerged was a clear link: A high use of Facebook was associated with lower ratings of wellbeing shortly thereafter. High Facebook use was also related to a decline in life satisfaction over the course of the study. On the other hand, frequent direct interactions with others increased (not decreased) feelings of wellbeing. A frequent explanation of these and similar results runs as follows: On Facebook, people are evaluating their normal everyday activities in comparison with the peak experiences and golden moments of others, and they naturally end up feeling inadequate and depressed. Kross et al refer to these evaluations as "damaging social comparisons" (page 4). Note that this longitudinal study is not an experimental design.

There is no manipulation, random assignment, or control of extraneous variables. 2 This interpretation was vividly supported by a September 2014 letter to Maclean's news magazine sent by a 16-year old from Toronto: Although the early days of picture-sharing and social media may have been about reconnecting with friends and expressing oneself, the practice has radically evolved into a fierce competition over which user has a better life in relation to his or her friends... It also allows users to deceive others into believing their "pictureperfect" lives are really as they seem. Kross et al's longitudinal design and statistical analyses allowed them to rule out the likelihood of reverse causationthat being anxious or lonely might subsequently cause people to spend more time on Facebook. It is still possible, however, that what researchers call a subject effect might have created this result. Specifically, if participants were aware of the belief that extensive Facebook use leads to negative feelings, this awareness might have biased their survey responses, consciously or unconsciously, and they might report more negativity simply because they were also reporting that they used Facebook frequently. To rule out the subject effect, participants must be unaware of the researchers' hypotheses. This was precisely the case in a recently published experiment by Adam Kramer and colleagues at Cornell University. This research group worked with Facebook to manipulate the content of the news feeds of 689,000 Facebook users. For a randomly-assigned group of these users, the researchers removed some posts that contained positive words, thus making the news feeds more negative.

For a second randomly-assigned group, they removed some posts that contained negative words, thus making the news feeds more positive. The study was conducted over a one-week period in January 2012 and users were not aware of these changes in content. The dependent variable was the emotional content of participants' own posts following the week in which the manipulation occurred. Compared to a control group, Kramer et al found that news feeds that were made more positive led to an increase in the positive words people posted in their own status updates. When news feeds were made more negative, the participants' number of Because this is not an experiment, the researchers are careful to address potential problems with internal validity. One such problem is the possibility of reverse causation. Because causes precede effects, however, the longitudinal nature of the design allows the research to rule out this alternative explanation. On the other hand, they cannot rule out the subject effect as a threat to internal validity. Manipulation and random assignment are key elements of an experimental design. The independent variable being manipulated is the Facebook news feed (more positive vs. more negative). Because the research took place in a natural environment, it would be classified as a field experiment (as opposed to a lab experiment). 3 negative words increased over that of a control group.

The researchers concluded: "the fact that people were more emotionally positive in response to positive emotion updates from their friends, stands in contrast to theories that suggest viewing positive posts by friends on Facebook may somehow affect us negatively, for example, via social comparison. In short, the results of this experiment appear to contradict the Kross et al interpretation of their findings. Why might this be so? It is possible, as argued earlier, that the subject effect invalidated the Kross et al results, and the Kramer et al results are more credible. Moreover, the huge difference in sample size might lead us to favour the Kramer result. Alternatively the inconsistency may have to do with the different measures of the dependent variable used by these studies. Kross and colleagues asked a straightforward question about participants' wellbeing; Kramer and colleagues counted the positive words in the participants' posts as a measure of their wellbeing. One might wonder whether a positive post in response to a friend's positive news really reflects wellbeing or whether it is just a competitive "keeping-up-with-the-Joneses" reaction. Negative words in response to a negative post might indicate an attempt at sympathy more so than a true decrease in wellbeing.

Given the popularity of Facebook and the questions raised by (or conflicting results of) such studies, the research on the impact of Facebook and other social networking sites will certainly continue. Beyond a discussion of the results, the Kramer et al article created great controversy. Critics declared that the study was unethicalthousands of Facebook users were made to be participants in research without their consent. Typically, research with human participants requires ethical approval by university ethics-review committees, and this is also normally a condition of publication in academic journals. Cornell University claimed that it was the Facebook corporation (now called Meta Platforms), and not the authors, who had collected the data; their researchers, therefore, were not culpable. Other defenders of the study mentioned that businesses with online information routinely carry out experiments like this for their own profit. The researchers are noting that their results are not similar to those found by other researchers.

This is not uncommon in the research process; an inconsistent result across studies is frequently the starting point of new research investigations. This is a debate about the ethics of experimental design. Typically, a study like this would not have passed an ethical review. 4 Yet it was disturbing to many readers that the participants' feelings of wellbeing were deliberately being influenced without their knowledge. Many Facebook users are teenagers and research on minors is rigorously reviewed by ethics committees since it is not deemed "minimal risk." It is generally believed that scientific researchers should be held to a higher standard of behaviour than businesses and marketers. Moreover, as one reader commented, perhaps corporations do engage in such studies, but this does not make them ethical.