Psych - Deception Flashcards
It was originally just said that experts are more attuned to kinematics + aim to disguise their actions…
However what is the extension of this?
Mann et al, 2007
Players try to deceive their opponents!
- it may not even be an advantage to be attuned to early kinematics… as they may be easier to fool!
Are deceptive actions frequent + effective - study rugby union carries
Wheeler, Askew + Sayers, 2010
- Sidesteps were bserved in 37% of all carries (frequent deception)
- When all tackle-breaks (missed tackle essentially) were analysed - 72% were a result of an evasive side-stepping strategy (effective)
Are deceptive actions frequent + effective - netball passes study
Fox et al, 2014
- 53% of all passes were preceded by an ‘offensive agility manoeuvre’ (frequent)
- 90% of which were side-steps (frequent)
- 68% of all deceptive manoeuvres were successful (effective)
Are high skilled performers more or less susceptible to deception - expert kick boxers study
Ripoll et al, 1995
Experts responded more frequently to ‘feints’ (47%) than did intermediates (30%) and novices (23%)
Are high skilled performers more or less susceptible to deception - experienced football players in 1v1 situations
Williams + Davids, 1998
Experienced players were faster + more accurate at responding to 1-on-1 situations (95% vs 77%)
- opposite results to kick boxer study
How can deception be examined experimentally…
What 3 different forms of evidence should researchers look for?
- Task performance is significantly below chance = not just guessing
- Significant decrease in performance across consecutive occlusion points (usually get better as temporal occlusion progresses / get worse in certain spatial occlusion scenarios)
- Poorer performance on trials containing deception
Evidence on disguise for england football penalty kick tackers
What key fact was noticed?
Jordet, 2009
GKs dived in the correct direction on 43% of the kicks (below chance so ae trying)
- however, they dived correctly for 60% of England kicks + saved 50% of them
- therefore England players are not good at disguising their actions
Temporal occlusion design study on skilled vs novice rugby players
Had to judge direction of ‘normal’ + sidestep actions occluded at 5 time points
What were their 2 hypothesis?
Jackson, Warren + Abernethy, 2006
A: Experts will be deceived more than novices (as are more attuned to early kinematics)
B: Experts will be deceived less than novices (as can better differentiate genuine + deceptive actions)
Temporal occlusion design study on skilled vs novice rugby players
Had to judge direction of ‘normal’ + sidestep actions occluded at 5 time points
Results?
Jackson, Warren + Abernethy, 2006
Hypothesis B was true
- Higher skilled players could better judge deceptive actions
- Video of national level players were more deceptive than recreational players
Temporal occlusion design study on skilled vs novice rugby players using stick men in VR
Again judging direction of sidesteps occluded at 4 time points
Results
Brault et al, 2012
Performance is really low on early occlusions for deceptive actions in novices + experts compared to non-deceptive movements in novices + experts
- rapid pickup of performance over just an extra 200ms of time (experts have earlier pickup at 100ms)
Multiple deceptive signals - foot displacement, head yaw + upper trunk yaw
- experts more attuned to COM = less distracted by peripheral distractions
Temporal Occlusion design study on skilled vs recreational rugby players in a lab where participants responded to video on a big screen directly in front
Results
Warren-West + Jackson, 2021
- Low skilled tended to focus more on the players’ head
- High skilled tended to focus more on hips + abdomen
- Focus was the same across the 2 windows (pre-deception and knowing you’ve been deceived)
This shows that the same cues being used to read deception are the ones deceiving you
How is exaggeration used by performers?
Use exaggeration during deceptive actions to create a larger response from their opponent
What locations did Brault et al (2012) identify as being exaggerated during deceptive actions?
Foot displacement, head yaw and upper trunk yaw
What did Pollick, Fidopiastis + Braden (2001) say about exaggerated serve actions?
They found that these exaggerated serve actions were recognised faster + more accurately
What’s the difference found in exaggerated, normal and deceptive penalty kicks?
Smeeton + Williams, 2012
Penalty kick direction was judged 20% more accurately in exaggerated than normal kicks and 20% worse in deceptive kicks
- When comparing deceptive + non-deceptive exaggerated = they’re pretty much identical in terms of visual gaze
How can contextual sources of info be deceptive?
Context can create bias expectations so may impair responses when actions are unexpected
Contextual sources info that can be deceptive - expectation bias
Study on GK saves during penalties
Explain briefly + results
Navia, Van der Kamp + Ruiz, 2013
Background: GK were given prior info about probabilities of what player is likely to do (e.g. 50% chance to left or 75:25 to the right)
Results:
- 50:50 group (no info) - saved 20% and dived correctly 50% of time = guessing
- 80:20 group - saved 25% and dived correctly 72%
- player goes to low-probability side - 16% saves and 31% dived correct
Contextual sources info that can be deceptive - expectation bias
Baseball batters having 3 preceding pitches (varying fast ball / normal in a sequence) and then got delivered 4th one
Results
Gray, 2002
- When the 4th ball was congruent with previous 3 throws (3 fast throws so expect a 4th) then swing timing error was very good
- But when 4th ball was incongruent - swing timing error was 2/3x greater
Contextual sources info that can be deceptive
How can misdirection + head fakes pose an issue?
Responses to passes with a ‘head fake’ are slower and more error prone
The head fake effect is larger when head fakes are rare (not used often)
Contextual sources info that can be deceptive - misdirection / head fakes
Temporal occlusion study design on male and female students (with no expertise) in judging basketball pass direction
Explain further
Guldenpenning, Jackson + Weigelt, 2023
Participants had to judge intended pass direction ‘as quickly and accurately as possible’
- were given player preference data (influence expectations) - (25/75, 50/50 75/25 in left/right)
- head fakes were on half of 288 trials
Contextual sources info that can be deceptive - misdirection / head fakes
Temporal occlusion study design on male and female students (with no expertise) in judging basketball pass direction
Results
Guldenpenning, Jackson + Weigelt, 2023
- when fake aligned with high probability value (look to 75% side) - then errors increased dramatically (amplified effect)
- slower RTs also
Integrating contextual + kinematic info:
What is Bayes Theory?
Predicts how contextual + kinematic info will be optimally weighted (Kording + Wolpert, 2006)
- takes initial prior expectation + says that will be updated based on evidence (kinematics etc)
- the prior (probability info) and likelihood (kinematic info) will be weighted accordingly to level of certainty
Integrating contextual + kinematic info - study on genuine vs disguised handball throws
Morphed videos creates different levels of certainty + action preference data (75/25 genuine/deceptive throws etc)
What did researchers expect?
Helm et al, 2020
If participant just goes with probability info…
- across all conditions (different ambiguity) then effect of probability should be same across 75% disguised throws, 50% and 25% (no integration at all)
If just focus on kinematics + ignore action preferences…
- expect rising pattern / curve
Integrating contextual + kinematic info - study on genuine vs disguised handball throws
Morphed videos creates different levels of certainty + action preference data (75/25 genuine/deceptive throws etc)
Results
Helm et al, 2020
Effect of prior was strongest at maximum uncertainty (50% disguised condition)
What is the signal detection theory?
Discriminating between 2 signals, while also measuring bias
Handy at taking results of trials and picking apart if…
- players are good at distinguishing between genuine and deceptive actions
- players are biased to judge action as deceptive or non-deceptive
Can SDT measure effect of context?
In lab study with football player on big screen + participant on pressure mat in front - judging whether player goes left or right
- given action preferences etc
Results
Jackson, Barton + Bishop, 2020
High skilled players accuracy fell from 95% (50/50) to 75% (83/17)
- 95% (almost perfect) when in ‘uncertain’
- 75% - couldn’t help but bite on deceptive actions when attached to high probability info
- response time did not change
Jackson, Barton + Bishop study (2020)
Does probability info bias player responses?
Every initial move has an outcome probability attached to it
- responding to it implies you judge the action to be ‘genuine’
- negative C (bias level) values imply a bias toward judging actions to be genuine
- bias to respond to high-probability side was stronger for deceptive actions than genuine actions
Low skilled players were more influenced by prior probability
Study on deceptive kicks (football penalties) changing eye gaze + approach angle
Asked experienced kicker to ‘produce an approach + kicking action as though intending to shoot from one side of goal before actually aiming to other’
Dicks, Davids + Button, 2012
Approach angle - 23 +/-5 vs 35 +/-5
Eye gaze - look toward / away from goal location
From 70 kicks…
- deceptive = 11% saved
- Non-deceptive = 56% saved
