“But That’s Just Twitter, Right?” Skepticism About the Use of Social Media Data in Forensic Evaluations

Ashley B. Batastini, University of Southern Mississippi
Michael J. Vitacco, Augusta University
Ashley C. T. Jones, University of Southern Mississippi
Riley M. Davis, University of Southern Mississippi

Perceived credibility of social media data (i.e., a Twitter post) was compared to more traditional collateral sources in criminal responsibility evaluations using independent samples of laypersons and forensic experts. Overall, results suggested greater skepticism toward social media relative to two other sources, particularly when information suggested a mental illness. Both samples, however, viewed the tweet as potentially useful. Notably, both studies were limited by the use of an experimental design that was intended to capture initial impressions rather than fully mimic standard assessment and courtroom processes. We advocate a cautious but open-minded approach when considering social media data as collateral.

Social media, forensic mental health assessment, criminal responsibility, NGRI

“Established standards in forensic psychology encourage evaluators to consider sources of collateral data when reaching psycho-legal opinions…The use of collateral data aims to improve the quality and credibility of forensic reports by not over-relying on one data source…A key aspect of the use of collateral data is its perceived utility and credibility…Although widely acknowledged as necessary to the development of appropriate opinions, there is a lack of research on how clinicians weigh and determine the credibility of data they encounter…The vast majority of early research and scholarship on the appropriate integration of collateral data into criminal responsibility reports was done prior to the proliferation of social media in the United States and around the world…the use of social media is routinely integrated into criminal forensic work, with criminal responsibility evaluations being the most common…despite the use of social media in forensic evaluations, examiners remain uncertain about the value of this information and its overall relevance…” (p.317-318).

“To examine whether information obtained from social media leads to differential judgements about its credibility compared with other more typical sources of collateral information used to inform psych-legal opinions, specifically criminal responsibility, this set of studies applied an experimental design to two separate samples…Study 1 represents the perceptions of a layperson audience…Study 2 aims to determine whether practicing forensic clinicians…view social media differently in terms of its credibility compared to other common collateral sources…we hypothesized that lay-persons and experts would perceive a defendant’s statements depicted as a mock Twitter post (or “tweet”) as less trustworthy, credible, and useful to the determination of criminal responsibility than the same statements documented in a clinical note or an excerpt from another examiner’s forensic mental health report…” (p.319).

“In each study, the content of the data source (i.e., what the defendant actually said) was also varied. Specifically, we assessed if perceptions of the source would change depending on whether the defendant’s statements were suggestive of psychotic symptoms or not. While we expected that content representative of a potential mental illness…would be viewed with greater reservation overall, we were primarily interested in the interaction between source content and type…We hypothesized that source trustworthiness, credibility, usefulness, and (for the expert sample only) weight assigned to the source would be rated the lowest when the defendant’s psychotic-like experiences were communicated via the tweet than when communicated to another professional or when the tweet contained a more normative experience” (p.319).

“Study 1 has potential implications for how jurors may perceive collateral information obtained from social media. Once a judge determines collateral information used by an expert was properly vetted and obtained, a jury will need to decide how much consideration to give the information when reaching their verdict…As suggested by the results of Study 1, jurors may discount information a clinician obtained from social media and consider it unreliable and not as probative to the presenting legal issue as other forms of collateral information. They could also decide that, if the social media is of limited value or was weighted incorrectly by the clinician, other aspects of the evaluation (including the forensic opinion) may also be less credible or trustworthy by association…” (p.328-329).

“Because forensic practitioners also appear to view social media data with greater uncertainty, there may be a propensity to prematurely disavow this information as relevant to their forensic findings. In contrast, the information found in the most traditional-looking clinic note and forensic report were more consistently held in higher regard…it should not be overlooked that all sources were essentially generated from the defendant’s self-report…A clinician who dismisses data generated from social media and misses pertinent information as a result may, in turn, create a vulnerability on cross examination. Instead…forensic practitioners are encouraged to carefully weigh the credibility and relevance of all collateral data, including those coming from social media. Failing to do so would appear to go against our professional principles concerning objectivity” (p.329).

“Ultimately, the jury will determine not only the credibility of the information obtained, but also the manner in which the clinician relied upon it. Therefore, it may be incumbent upon attorneys from the retaining party to structure their questioning such that expert witnesses are able to reiterate the credibility of data derived from social networking sites, including how it was obtained and how it was used by the expert. Conversely, opposing counsel could attempt to discredit the information on cross-examination, increasing the risk that jurors will inappropriately weigh the evidence, and forensic practitioners should be prepared for this possibility. Judges may also opt to educate jurors about the use of social media data in specific instructions prior to deliberates…” (p.329).

“In fact, it is conceivable that spontaneously generated statements found on social media platforms could be more accurate and relevant to criminal responsibility when compared to a potentially scripted conversation with a medical and/or mental health professional. In many instances, perhaps especially in criminal responsibility evaluations, data from social media may prove more relevant as they have not been filtered through the lens of another individual or tarnished by the passage of time. Future research will continue to explore potential uses and misuses of social media data by forensic practitioners in answering psycholegal questions” (p.329).

“As the reliance on social networking sites as a collateral sources increases, there is an absence of professional guidance on how to properly integrate online data of this nature into forensic opinions, especially as such data can be difficult to interpret…Forensic practitioners are strongly encouraged to familiarize themselves with extant literature and relevant case law (e.g., Griffin v. State of Maryland, 2011) to remain current with what is undoubtedly going to continue to be a rapidly evolving practice…Prior to testimony, it may be useful for experts to discuss the literature on social media perceptions with retaining counsel and recommend they be given an opportunity to address such concerns during direct examination” (p.329-330).

“…In addition to using methodological designs that increase external validity (e.g., mimicking courtroom procedures more closely, informing participants that all data was provided through discovery…), future work in this area should explore perceptions of credibility across varying social media platforms and digital communication formats (e.g. video vs. text-based)…Relatedly, additional research should directly explore the specific concerns laypersons and experts have when it comes to the veracity of data derived from social media through the use of qualitative data collection methods; skepticism could be rooted more so in authorship, potential for impression management, or contextual concerns…” (p.330).

As always, please join the discussion below if you have thoughts or comments to add!