Twitter is increasingly investigated as a means of detecting mental health status, including depression and suicidality, in the population. However, validated and reliable methods are not yet fully established. This study aimed to examine whether the level of concern for a suicide-related post on Twitter could be determined based solely on the content of the post, as judged by human coders and then replicated by machine learning. From 18th February 2014 to 23rd April 2014, Twitter was monitored for a series of suicide-related phrases and terms using the public Application Program Interface (API). Matching tweets were stored in a data annotation tool developed by the Commonwealth Scientific and Industrial Research Organisation (CSIRO). During this time, 14,701 suicide-related tweets were collected: 14% were randomly (n=2000) selected and divided into two equal sets (Set A and B) for coding by human researchers. Overall, 14% of suicide-related tweets were classified as ‘strongly concerning’, with the majority coded as ‘possibly concerning’ (56%) and the remainder (29%) considered ‘safe to ignore’. The overall agreement rate among the human coders was 76% (average κ=0.55). Machine learning processes were subsequently applied to assess whether a ‘strongly concerning’ tweet could be identified automatically. The computer classifier correctly identified 80% of ‘strongly concerning’ tweets and showed increasing gains in accuracy; however, future improvements are necessary as a plateau was not reached as the amount of data increased. The current study demonstrated that it is possible to distinguish the level of concern among suicide-related tweets, using both human coders and an automatic machine classifier. Importantly, the machine classifier replicated the accuracy of the human coders. The findings confirmed that Twitter is used by individuals to express suicidality and that such posts evoked a level of concern that warranted further investigation. However, the predictive power for actual suicidal behaviour is not yet known and the findings do not directly identify targets for intervention
Currently, little is known about the prevalence of mental disorders in athletes. This study aims to investigate Australian elite athletes' symptoms of general psychological distress and common mental disorders. A cross-sectional survey design was employed to assess self-reported symptom prevalence. A total of 224 elite athletes (118 female, 106 male) from national sporting organisations in Australia were administered a self-report internet-based survey comprising measures of demographic status and mental health symptoms. Overall, 46.4% of athletes were experiencing symptoms of at least one of the mental health problems assessed. Percentages meeting criteria for mental disorders were similar to previous epidemiological studies of both international athlete and community samples: depression (27.2%), eating disorder (22.8%), general psychological distress (16.5%), social anxiety (14.7%), generalised anxiety disorder (7.1%), and panic disorder (4.5%). Injured athletes had higher levels of both symptoms of depression (t=3.23, p=.001) and generalised anxiety disorder (t=2.26, p=.025). The level of symptoms of mental health problems reported by elite athletes appears similar to that observed in the community. However, caution must be exercised in interpreting the findings, as possible demographic differences between athletes and comparison population datasets may exist. Furthermore, self-selection of respondents in the present study may have reduced the representativeness of the sample and the validity of the comparisons. Athletes, particularly those currently injured, should be well-supported to seek help for mental disorders through access to mental health professionals.
Recruiting participants is a challenge for many health, medical and psychosocial research projects. One tool more frequently being used to improve recruitment is the social networking website Facebook. A systematic review was conducted to identify studies that have used Facebook to recruit participants of all ages, to any psychosocial, health or medical research. 110 unique studies that used Facebook as a recruitment source were included in the review. The majority of studies used a cross-sectional design (80%) and addressed a physical health or disease issue (57%). Half (49%) of the included studies reported specific details of the Facebook recruitment process. Researchers paid between $1.36 and $110 per completing participants (Mean = $17.48, SD = $23.06). Among studies that examined the representativeness of their sample, the majority concluded (86%) their Facebook-recruited samples were similarly representative of samples recruited via traditional methods. These results indicate that Facebook is an effective and cost-efficient recruitment method. Researchers should consider their target group, advertisement wording, offering incentives and no-cost methods of recruitment when considering Facebook as a recruitment source. It is hoped this review will assist researchers to make decisions regarding the use of Facebook as a recruitment tool in future research.