Discussion

Source:

Chapter 9 – Writing for Impact: How to Prepare a Journal Article

Andrew M.Ibrahim, Justin B.Dimick

https://www.sciencedirect.com/science/article/pii/B9780128099698000097?via%3Dihub

This section connects  the previous parts: introductions, methods and results. It consists of several paragraphs as follows.

  1. Paragraph that summarize the findings: tell the reader what the study was about and what the results were.
  2. Paragraphs that  put all results into the context. You may start with summarizing previous similar studies and add your findings that may agree and extend the prior works. Or, your work may  challenge the previous works.
  3. Paragraph that describe limitations of your study. These may invalidate your result. The limitation could be: random error, systematic bias and confounding factor. You should try to explain how do you mitigate these limitations.
  4. Ending Paragraph where you discuss implication (s) of your study. You have to avoid saying: “more research is needed”. It may de-value your contribution. Instead, you express your deep-thought of the matter and your intension to advance the study further. It is suggested that you ask a senior author to help you in this paragraph.

Example

4.1 Irrelevant as high salient experiments

There are a number of points to consider with regard to the Irrelevants as high salient countermeasure. In particular, we did see a change in the Irrelevant ERP and recall/recognition of Irrelevants in experiments three and four, when participants were attempting the Irrelevants as high salient countermeasure. However, this change in the electrophysiological and behavioural pattern was not sufficient to confound the Fringe/P3-Rapid method. Furthermore, the Irrelevants’ P3 pattern remained very different to, even, the Fake pattern. This is important, since a likely consequence of artificially elevating the salience of an Irrelevant is to make it a task-prescribed target, in much the same way as the Fake.

Thus, it would seem that, when participants “see” the Irrelevant, it is sufficiently late in the experiment that any P3 present in those (late) trials is “watered down” when averaged against the earlier P3-absent trials. Note, such late detection of Irrelevants would lead to an increase in recall and recognition of these Irrelevants (which we see in a limited number of cases); after all, our memory tests do take place after all trials have completed. But, such an increase in memory could only reflect very late identification of the Irrelevant.

4.2 Novelty of Our Approach

This is the first study demonstrating that presenting stimuli on the fringe of awareness impairs perception of non-salient items, hindering countermeasure use. Our approach differs from a related proposal by Lui and Rosenfeld, whose data supported the hypothesis that lie-related stimuli, which are presented subliminally, may differentially affect ERP patterns of subsequent, supraliminal stimuli [18]. In contrast, our hypothesis requires presentation of stimuli at near-subliminal speeds (on the fringe of awareness), allowing perception of salient stimuli, including items which carry concealed knowledge. Specifically, the ERP comparison we are making is between a conscious percept (when the Probe “breaks through” into awareness), and the absence of such a conscious experience (as arises for Irrelevants). The large ERP differences we observe between Probe and Irrelevant seem to reflect this—presence vs. absence of a conscious experience. Electrophysiological responses formed in subliminal priming experiments do not reflect this distinction—the objective in subliminal priming experiments is to render all primes subliminal—whether salient (as a Probe is) or non-salient (as an Irrelevant is). Thus, subliminal priming experiments do not induce distinct conscious versus non-conscious states of experience between different classes of prime. Really, the effectiveness of the Fringe/P3-Rapid deception detector rests upon the perceptual regime in which RSVP places the brain; that is, stimuli are presented such that only a small subset of them can be consciously perceived, and, critically, the brain selects those to perceive on the basis of their salience. In other words, only salient stimuli break into consciousness and set-up the pronounced electrical response we see in the P3a and P3b.

This study demonstrated high accuracies in the classification of deceivers and non-deceivers based on ERP data alone, even when countermeasures were applied. Previous studies have indeed demonstrated that one’s own name, one of the most over-rehearsed stimuli, can elicit large electrophysiological responses, particularly in frontal regions [33]. The P3a (or novelty P3) was also elicited in comatose patients, by using their own name (as an auditory stimulus) [34]. More generally, the P3a is often elicited when a participant’s own name is presented as a task-irrelevant stimulus [35][36]. Because of these precedents, we have used the term P3a to identify the early fronto-central oscillations we obtained, although we acknowledge that our P3a patterns differ somewhat from those typical of oddball-type experiments (which are characterised by a slightly later latency and rapid habituation [37]).

The use of first names as stimuli may explain the very high hit rates we obtained, thanks to the strength of our P3a component. Nevertheless, the P3b pattern should be reliable across stimulus types and even if our approach were to show its largest effects in identity deception, it would still be of interest: countermeasure-resistant identity deception is a valuable tool for forensic science (including detection of simulated amnesia). Moreover, P3-based deception detection systems based on the classical oddball paradigm have recently been demonstrated to be vulnerable to countermeasures derived from directed forgetting techniques [38]. The efficacy of such techniques applied to our paradigm remains to be verified.

Countering Countermeasures: Detecting Identity Lies by Detecting Conscious Breakthrough

Howard Bowman, Marco Filetti, Abdulmajeed Alsufyani, Dirk Janssen, and Li Su

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0090595