Below you find a very brief summary about the Concealed Information Test that is implemented in this app. To properly understand the test, please read the online
documentation.
The standard and enhanced versions aim to reveal whether or not a certain information detail – the "probe" – is known to the tested person. The "irrelevant" and "target" items must be similar to the probe, and indistinguishable
for a person who does not know the relevance of the probe (e.g. the probe is a stolen suitcase, and the irrelevant and targets are other suitcases). In this app, the input requires no designated "irrelevant" items: all items are potential
probes, and the real probe can be given as any of these items. In the end, each of the items (except the target) will be automatically evaluated as potential probes (in each case with the given other four items, excluding the target, being
irrelevants).
The enhanced version is much more effective than the standard; the latter is only for experimental purposes. The no-target version is a tentative approach in development to make the test applicable in cases where the
probe is actually known to the participant.
At the end of the test there are no instructions or button, but only the text "Test completed." (in the selected language). This is to prevent subjects to see the results without permission. To show results, swipe right on the
text.
Demo
For demonstration's sake, take a situation in which you want to reveal whether a person is aware of a certain date. For example, someone is suspected to be concealing the knowledge of an upcoming terrorist attack on May 9. In this
case, five
different, additional dates should be chosen randomly. For example, June 14, December 5, August 25, February 12, and October 23. From among these, one should be chosen randomly as target, for example, August
25. These should be filled in
accordingly on the starting page: AUG 25 as Target, and the other dates as Probes (1-5). (In this case, the real "probe" is MAY 09, but the app automatically evaluates all items as potential probes.) The Subject ID could be "CIT_demo_suspect_01".
All these data can be filled in automatically for a demonstration using the following button:
Demo data loaded.
If you would like to do a pilot test to see that the method truly detects a relevant detail, one good way is to enter your own personal name (e.g. family name) as one probe and, as target and other probes, enter other, randomly chosen names.
If you then
complete the test in the enhanced version with, say, two blocks, it is very likely that you will be detected. (Personal names are typically highly personally relevant, and thus appear highly salient in the task.) For a control test, to show
a no effect (simulating innocence), you may enter random names for all items. Note that, in this case, the person tested should not know which item is the presumed real probe item! The mere knowledge of the
relevance of this item (as the probe to be tested) can cause different responding.
Results
Path to files on the device(tap to show)(tap to hide)
{{citP.path}}
{{citres.v.subj_id}} ({{citres.v.date}})
Summary CIT results for {{citP.cit_results.subj_id}}
Overall accuracy rates (number of correct responses per number of all trials): {{citP.cit_results.ar_overall}}.
Description of variables(tap to show)(tap to hide)
Response times (for all correct responses, in milliseconds): probe RT mean minus irrelevant RT mean (RTP vs I), probe RT mean and standard deviation (RTP (SD)), irrelevant RT mean and standard
deviation (RTI (SD)),
uncorrected Cohen's d between all probe and all irrelevant RTs (dCIT). Accuracy rates (ratio of correct responses compared to incorrect and too slow responses, in percentage): probe accuracy rate minus irrelevant
accuracy
rate (ARP vs I), probe accuracy rate (ARP), irrelevant accuracy rate (ARI). Responses with RT below 150 ms are excluded from all RT and AR analysis, except for the overall accuracy rate
calculation. Trials (i.e., responses) from practice rounds are not included in any of the statistics above.
Evaluation using dCIT(tap to show)(tap to hide)
The dCIT measure might be used to evaluate whether or not a given probe was recognized. The dCIT typically falls between around −0.3 and 0.8, where a larger number always indicates a larger likelihood
of
the
recognition of the given probe. As of yet, there is no established optimal cut-off value for the evaluation, but, based on a rough approximation in view of previous results, the following table depicts some possible evaluative labels for
given boundaries.
dCIT > 0.4
strong indication of recognition
dCIT > 0.3 and dCIT
<= 0.4
fair indication of recognition
dCIT > 0.1 and dCIT
<= 0.3
weak indication of recognition
dCIT > 0 and dCIT
<= 0.1
indeterminate
dCIT > −0.1 and dCIT
<= 0
weak indication of non-recognition
dCIT
<= −0.1
fair indication of non-recognition
(There is no strong indication of innocence: a very fast probe response is no more expected in case of innocence than in case of guilt.)