Deception Detection In Non Verbals, Linguistics And Data.

Thursday, June 30, 2016

Judging Competence + Success From A Face.

Previous studies have shown that personality traits such as competence and trustworthiness can be reliably judged from a face (Hess, Adams, & Kleck, 2005). Similar studies showed a 70% correlation to predicting election results in the U.S based on looking at photo's and scoring the same traits.

However, can you tell which CEO's are most successful from their face?

                                                                                                         credit-- inc.com


Undergraduates were asked to judge CEO's likability and competence based on a series of photo's in a study by Nicholas Rule and Nalini Ambadi:

http://psych.utoronto.ca/users/rule/pubs/2008/Rule&Ambady(2008_PsychSci).pdf

It turns out that leaders that scored the highest also ran the most successful companies. Nicholas Rule says, "These findings suggest that naive judgments may provide more accurate assessments of individuals than well-informed judgments can."

First impressions are critical, because we do judge a book by it's cover. Which brings me to Bill Shortens demeanor when interviewed on TV.

I have noticed in the last week that Shorten appears to be going into "sad mode" as he begins to speak. This morning while being interviewed on channel 9 with one day to the election, Bill Shorten's face visibly changed at the moment he began to speak. His eyebrows over the nose went went together and up, his top eyelids dropped slightly, in a classic sad expression.


Whether this is to invoke underdog sympathy, or portray a large burden being placed on his shoulders, it appears to be a conscious attempt because he visibly changes as he begins the interview.

The world's pioneer and authority on facial recognition, Dr. Paul Ekman in his book Emotions Revealed says,

"The eyebrows are very important, highly reliable signs of sadness."
And when talking about actors he says, "It makes them seem more empathetic, warm and kind, but that may not be a true reflection of what they are feeling."

I wanted to test whether Bill Shorten is using this as part of his persona. I downloaded all his speeches for the month of June as well as Malcolm Turnbull to use as a comparison.

I ran the speeches through the psychological text analysis tool LIWC from James Pennebaker from the University of Texas to categorise dozens and dozens of words related to sadness from all the speeches.




Compared to Turnbull, Shorten uses more sad language, using more words like lose, missing, tragic, deprived.



This persona may be part of the political spin developed by Shorten and the Labor campaign to invoke empathy, but portraying trustworthiness and competence has been shown to be a better combination.
SHARE:

Monday, June 27, 2016

Election Word Watching - Who's Most Deceptive?

I downloaded 28 speeches from Malcolm Turnbull and Bill Shorten (14 each) for the month of June. This includes scripted and unscripted Q+A sessions. I ran all the speeches through the text analysis software, which used 4500  words in 80 categories to to analyse what was said and to get deeper into the state of mind and intent of the leaders.

The most frequent words used by both leaders shown here as a word cloud. The larger the word, the more often it was used.


Above: Bill Shorten word cloud.

                                                  Above: Malcolm Turnbull word cloud

While both leaders mention the other side, Shorten stands out with how often he uses the word "Turnbull", it's nearly as large as Labor.

Next we look at Equivocation or Hedge words (I believe, think, might, should could etc) which reduces commitment and can "indicate a less positive experience or an unwillingness to communicate information". (Wiener and Mehrabian 1968)

I'll also look at Negations, words like no, not and all contractions of not. Equivocation and Negation were pinpointed by Susan Adams of the FBI as the most indicative indicators of deception in her paper Indicators Of Veracity And Deception: An Analysis Of Written Statements Made To Police - 2006.




The results are highly significant at below the 1% level. Bill Shorten uses far more Equivocation and Negation compared to Malcolm Turnbull. This is a red flag in deception, but it is also high in losing political parties in the 100 year election speech analysis I did. Winners of elections used less, losers used more of these words.





Tracking what drives the election leaders, we break down affiliation, achievement and power indicators in speech. People high in affiliation are concerned with relationships and close allies. Both leaders are similar in affliation and also achievement orientation.

The power indicator -- how driven is the individual to control, status and prestige. David Winter argued in his analysis of U.S presidents that the degree of power was an indicator of political effectiveness, but I have found it to be a negative indicator in Australian elections - the public don't seem to vote for power language used in elections.

Looking at words relating to anger, Turnbull scores higher (more angry). This is a negative indicator and is also a negative indicator in the German election prediction using twitter to determine public sentiment that I mentioned in a previous post.

Both leaders are very similar on indicators of Analytical thinking, both are similar in Risk and Reward indicators, and both are similar in Authenticity.

Shorten uses far more "I" pronouns which tends to be "more personal, but more insecure." (LIWC analysis by psychologist James Pennebaker).





Both leaders are similar in focus on the future and the past events, but Bill Shorten is far more likely to focus on the present. There doesn't appear to be that much of a difference on the graph, but it is highly significant on the statistical analysis.

Pennebaker says, "People oriented towards the present are thinking about current events that are psychologically close. Present-focused people tend to be more neurotic, depressed, and pessimistic than either past- of future-oriented people. "

In total Malcolm Turnbull only has 3 categories of words out of 80 that are more statistically excessive than Bill Shorten, whereas Shorten has 12 categories.

Shorten has a problem with language compared to Malcolm Turnbull (this only relates when compared to each other) - he is perceived as more negative and is flagged as more deceptive.

In closing, the sincerity problem Shorten has is really shown in these pictures (gifs) below.
He is delivering a speech where he says, "I believe..." then looks down at his notes to remind himself what he believes in. This really is indicative of poor preparation or going into "auto pilot" during a speech.

But just when you think he was tired or having a bad day, he does it again during a second important speech about his asylum seeker/refugee policy!







This looks to all the world that Bill Shorten doesn't know what he believes.

Wrapping this up with the latest Befair prices on the election, Labor is diving further, with Liberal at $1.13, almost 89% chance of winning with less than a week to go.


SHARE:

Sunday, June 26, 2016

Power handshakes and other election non verbals.

Handshakes are often our first and sometimes only point of contact we have with another person. How we do it, affects how we are perceived. The handshake has the power to leave an impression, so it is interesting watching how the election leaders approach this.



Turnbull tends to pull in Shorten when they shake hands to establish dominance. This can appear very negative if it's exaggerated as with Tony Abbott's aggressive hand jousting exhibition with Kevin Rudd.


Turnbull appears more subtle--



Although with Prince Charles, Malcolm Turnbull made no attempt at a power handshake:


Possibly the very worst handshake you could do, and one that almost always leaves a negative impression is referred to as the "Politicians Handshake" and and involves the left hand covering the handshake in a two handed gesture, in an attempt to appear more friendly.


As with Shorten's jacket-off-and -rolled-up-sleeves approach, it's manufactured to make him appear as friendly, one of the people, let's get to work look.

During the election campaign, Malcolm Turnbull has come across as more confident, competent and relaxed then Bill Shorten. Whereas Turnbull stands up straight and hold his head up, Shorten tucks his chin in, a sign of discomfort.


AS a small child, if you smelt something disgusting, you wrinkled you nose and pursed your lips. Thirty years later, if you read a contract or see something you don't like, you purse your lips again. The lip pursing is normally only a brief moment, but it's an extremely reliable cue for discomfort.

When there is tension or stress, we have tension in our mouth area, which results in the lips being "sucked" in, and extreme discomfort or stress can make the lips disappear completely. Look around the airport next time you fly, and watch people when a flight is cancelled. Again, this is a very reliable indicator of stress and discomfort.



Shorten hasn't been coached to control overt discomfort and stress cues because even the Press have picked up on this, exhibiting his most extreme displays:


http://www.smh.com.au/federal-politics/political-opinion/alex-ellinghausen-photographing-malcolm-turnbull-tony-abbott-and-bill-shorten-20151210-glkboq.html

During interviews and debates, Malcolm Turnbull has his non verbals mostly under control, ensuring that he looks confident. Confidence is correlated to perceived competence in studies, so it is critical that a leader appear confident.

He speaks with large open gestures showing the palms of his hands, he stands straight, he tends not to point, which is highly offensive for many people. Turnbull's large open gestures are contrasted with Shortens more closed approach....hands closer to the body and sometimes pointing.



Millions of years ago, when we didn't like what we saw or were intimidated, we would run away. Today in a business setting this translates to us leaning away from someone who says something we don't agree with. We lean back in our chair, are lean away, and in a more extreme case we turn our body away from what we don't like or agree with.

This is evident when you watch political debates, but it is another problem that Bill Shorten has. He turns away slightly or looks with a sideways glance. Covering or turning away, or "ventral denial" shows discomfort with what is being said or asked. An easy way to see this is to watch which direction the feet point. The feet point to where the body wants to go.

If you talk to a client and their foot points to the exit door, they need to go. Jury consultant Jo Ellen Demetrius cites a study of jury members --when jury members don't like a witness, they face the witness but their feet point to the exit door.

So watch where the body (use the belly button as a directional indicator) points, and be aware of the more subtle version of feet pointing away.


Bill Shorten is not fairing well with his display of non verbals during this election campaign. I'll look at the verbals of the leaders in the next post.

SHARE:

Thursday, June 23, 2016

"Why Most Published Research Findings Are False."

Why False Positives are the downfall of most research papers.

The headline comes from this academic paper --
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1182327/

It's based around a massive statistical problem at the moment. Simply put, the problem is this - because computers and software have become so powerful,  multiple testing is becoming the norm.

So whereas many years ago one experiment would have been carefully considered and a hypothesis would have been formulated, then the testing would determine whether this hypothesis was correct or not at say the 95% level (the most common).

Above graph from paper showing brain scan data grossly exaggerated because of incorrect statistical tests and controls.


This means that luck and False Positives (thinking you've found something when you haven't) would be at around the 5 % level and you could be fairly sure you had a positive result at the 95% level.

Nowadays, by running 20, or 50 or 100 tests and looking for an effect after the fact, it creates a massively biased data set. It's not 5% False Positives anymore, it's more like 35%-40. (in brain scans).

You are now guaranteed to to have lots and lots of positive effects that are due to luck alone, It's called data steering and many other terms. The bias is so large you don't have a result even if you think you do.

In neuroscience and brain scans, they are testing thousands of voxels at a time. The errors become magnifies the more tests that are being run, and they realise they have serious analysis problems with fMRI scan analysis --

http://reproducibility.stanford.edu/big-problems-for-common-fmri-thresholding-methods/




This helps to explain the fact that studies are unable to be replicated in many/most cases. Nowadays, only 18% of Pharma tests pass Phase 2 stage, only 50% pass Phase 3 stage.

It's been estimated that since 2004, only 7% of studies have accounted for this multiple testing error. So all studies from drug trials to economics that do not use some form of Multiplicity Control (compensating for that fact that multiple testing has been done) are useless!

So what does this have to do with word + deception analysis? Everything. Many variable are considered, and some kind of multiplicity control to compensate for this is vital.

I emailed an Italian professor, Livio Finos at the University of Padova who wrote the MatLab code for this book and from which I have based my statistical tests-


I checked with him if I had the procedure correct that he advocated in his paper with a new method of Multiplicity Control -- http://link.springer.com/article/10.1007%2FBF02741320

He confirmed by email that I had the correct procedure. I then hired a Ukrainian freelance programmer Victor, and after a few days of back and forth email communication, it culminated in a 4 hour skype session in the middle of the night and I managed to get the MatLab code written and tested. This means I can correct for multiple tests with a new efficient process besides the industry standard FDR approach, as well as using Non Parametric Permutation Tests for all analysis, so the results I get are reliable.

SHARE:

Turnbull, Shorten + 100 Years Of Election Speeches

How does Malcolm Turnbull's election speech and Bill Shorten's Budget response compare to the last 100 years of election speeches in Australia?

To find out, I downloaded election speeches from 1903 from:
http://electionspeeches.moadoph.gov.au/explore



I cleaned up the data, removing things like [APPLAUSE], and used a psychological text analysis tool called LIWC (Linguistic Inquiry And Word Count) developed by James Pennebaker at the University of Texas and which has been verified and used in over 6000 articles and studies on Google Scholar.

The program looks at about four and a half thousand words in eighty categories. These categories tapped emotional and cognitive dimensions of the speaker, revealing things like state of mind, the tone and how analytic someone was, as well as keeping track of the almost "invisible" function words used in speech.



Custom MatLab code was used to find highly significant groups of words that was able to separate the winners from the losers over all the elections.

This in itself was amazing to me, I wasn't sure whether there would be a clean separation between what the winners of the last 100 years of elections said, and what the losers said.

It turns out election winners tend to use certain language, as do the losers. Comparing Malcolm Turnbull's election speech to this "model" placed him in the winner group, while placing Bill Shorten in the losers group by a long margin, hence the prediction of Turnbull winning by about 85-90%.


Some of these categories are positive (which winners have more of) and some are negative (which typically losers have more of). So for instance, the text analysis program has groups of words relating to power-awareness - this captures the degree to which people use words such as command, boss, victim and defeat. This measure to the degree the awareness of the power they have. In Australian politics, this equates to less being better, it tended to be higher in the losers group.

Turnbull comes across higher on the Authentic algorithm (capturing cognitive complexity and relatively low rates of negative emotion) and also his Tone is higher, both being positive.

On the downside for Turnbull has greater use of the word They (from the last post) which is negative.



Shorten has a problem with far too much equivocation, this is hedging language which reduces commitment and allows one to minimise what has been said if it turns out to be wrong at a future date. Words like I believe, I think, I thought, suppose and so on. He also uses too many negations such as couldn't, should't and wouldn't. (which is an indicator in deception or spin if it's not in response to a specific question).

This analysis was Turnbull against ALL the election winners, then against all the election losers; the same was done with Shorten. It compared them to the model - how well did they fit in the previous election winners group, how well did they fit in the losers group.

It did not compare Turnbull directly to Shorten with their speech, I will be doing that next post.
SHARE:

Sunday, June 19, 2016

The politics of WE and THEY.

The Secret Life Of Pronouns by James Pennebaker used statistical analysis and research to show that pronouns such as I and we and they as well as the small functions words, which are essentially "invisible" to us in day to day communication, actually revealed our feelings, our state of mind and were close to being a fingerprint for each individual.


Credit: Jane Fennessy/Blue Vapours

The way we wrote revealed a lot of psychological information about us as well as being unique to such a degree, that anonymous text could attributed to an unknown writer.

J.K Rowling was unmasked (and admitted) to being "first" time writer Robert Galbraith, and was detected solely on her use of language, in particular the small pronouns and function words.


So what does this have to do with politics? Words, in particular pronouns, are vital tools of the politician. During a speech or interview they attempt to present themselves in a positive light, while presenting negative aspects of their opponents, and pronouns in particular are used for this purpose.

One way of doing this is the WE (or us) against THEM dichotomy. Vote for us and you'll be better off, vote for them and you'll be worse off.

And with politics becoming more personality driven as major political party become less ideological, politicians present different identities to different voter groups in an attempt to relate to diverse groups. A key persuasion principle is that people like people that are similar to them.

The meaning of WE is that of group membership, it's the rest of the group that includes me. Politicians use of the pronoun WE has several connotations - to talk on behalf of their party, to reduce their individual responsibility or to to include or exclude listeners from a group. This makes WE a very useful linguistic tool for the politician.

"WE as a nation stand tall..." refers to all of us.
"WE have been able to build ever closer links...." now refers to the party.
"WE are a true and trusted friend..... " now refers back to the all of us.
"WE are called upon to do so...." is ambiguous, but refers to a political event.
"What do WE know?" refers to all Australians.
"WE know that the Liberals are in danger of blowing out the budget......." now becomes the political party.

So in using WE, there is a continual shift between all of us (all Australians) and the party, and the listener can include themselves in that group or not.

THEY is similar as a distancing tool: "THEY are not able to balance the budget", "THEY are worried...", "THEY will say......"

THEY can include the Australian people or the party, so a politician saying "THEY won't be fooled by...." is distancing himself from that claim by attributing that premise to all Australian people.

So WE and THEY involve distancing or including, allowing the listener to be part of that group or not, and even as an ambiguous reference allowing the politician to appeal to a wider group, and allowing the comment to be "softened" or negated at a later date.

The primary goal is to always represent oneself in a positive light, and WE and THEY are very useful for that.
SHARE:

Wednesday, June 15, 2016

Betfair Odds on Liberals Diving....

A five cents dive on Betfair for the Liberals in 24 hours puts them at 86% of winning.

The markets are nearly in the middle of my estimate from 5 weeks ago of 85-90% probability in Liberals favour.

Yet just this weekend media reports have them "neck and neck".
I think not.


SHARE:

Tuesday, June 14, 2016

Elections and Deceptions

Election mode in Australia and the U.S means electoral spin and truth stretching, although in Trump's case it was stretched and broken long ago -- Pulitzer winning web site Politfact.com  and their truth-o-meter have tracked and verified Trump's election statements as 70% false or lies!

Politics is becoming less about policies and political messages and more personality driven as the major parties begin to look more and more similar.




Studies from the U.S show over 70% accuracy in predicting election outcomes by showing voters photographs of politicians and asking them who looks more competent, and who looks more trustworthy.

With a more cynical electorate, politicians try to appeal to a wide range of people, to be everything to everyone, and so represent themselves with varying degrees of spin - a form of deception.

Politicians don't answer questions that would make them look bad or alienate part of their electorate, and so use equivocation and hedging to minimise the impact of what they are saying, even ignoring the question (ignoring the question in a political interview happened around 40% of the time, UK, Peter Bull, Claptrap And Ambiguities).

So words from speeches and interviews can be used to track spin and honesty among politicians. And just like in FBI Statement Analysis, we tend to see discomfort and stress in the communication leakage of politicians who are stretching the truth. We see it in the "I" pronoun usage being reduced as they distance themselves from what is being said, we see it in an increase of words conveying negative emotions (Pennebaker, Uni Texas), and see it with an increase of action words to keep the story moving along.







Words have also been used in election  and voter sentiment prediction, such as tens of thousands of twitter feeds being used to predict the German election outcome based on voter sentiment--




Jeremy Frimer also tracked voter sentiment with what he calls Prosocial language, plotting a direct correlation to language used in Congress and voter sentiment plunging to an all time low of 9% satisfaction with Congress.




I have analysed election speeches from Malcolm Turnbull and Bill Shorten, as well as all the election speeches of every Prime Minister and Opposition Leader in Australia going back to 1903 using psychological text analysis software and custom MatLab code to determine if it was possible to create a model on linguistics alone, on what is said in speeches and interviews....to separate the winners from the losers.

It turns out that there are 9 categories or "bundles" of words that Australian election winners use more of, and election losers use less of, and vice versa. This is significant at at the 5% level.

This allows for a fairly easy creation and testing of a linguistic model for winning Australian Elections.

So for example, words that relate to negative emotions (nearly 400 words selected by psychologists are in this group) are highly significant, and using them reveals conscious and even subconscious negative emotions about what is being said. This type of analysis has also been used in criminal statement analysis.

I ran my model over Malcolm Turnbull's speeches as well as Bill Shorten and gave Turnbull an 85-90% probability of winning. This was 4 weeks ago when the Befair odds on Liberal were $1.45. I was hoping they would lengthen for the Liberals, but only Shorten has lengthened!

Alas, Turnbull is at unbackable odds of $1.22 at the moment, translating into about 80% chance of winning, so the Betfair market is moving closer to my assessment.


This is despite the media hyping up the chances of Labor, painting the election as "too close to call" about a week ago. The significance of the Betfair market (and most betting markets) has been well known for quite a while as far more accurate than polls.

The lower take on winnings from Betfair and zero longshot/favourite bias means a more accurate barometer than bookmakers, media and polls. See for example http://researchdmr.com/RothschildPOQ2009.pdf





Of course, it's not just politicians stretching the truth during a campaign, it's media too. Each have an agenda to push.

Language is never neutral. The same words can mean different things to different people because we all carry our own internal dictionary. When we listen to someone talk, we assume things based on our interpretation of what they are saying. Words create pictures in our mind.

That's why the Bush campaign used "Tax Relief" instead of "Tax Cuts"; the visual minds-eye message is much more powerful because you are now giving people a relief from a burden. And so "Drilling for Oil" becomes "Energy Exploration", "Gambling" becomes "Gaming" (you don't think of a loser in a raincoat with a crushed up form guide anymore, gaming is something new and exciting with family-friendly venues).

Whenever euphemisms are used, such as the military use of "collateral damage" or the media love of the legal euphemism "execute" for a wanton murder, an agenda is being promoted.

Words are the toolbox of persuasion. Language experts are on both sides of politics, advising what words to use, which phrases poll higher. Experts like Frank Lund for the Republicans and George Lakoff for the Democrats.




Both of these experts have written books which can be bought on Amazon, and show the current state of the art as far as political linguistics goes.

If you accept a word and use it, you accept the frame (as Lakoff calls it) that comes with that word.

Different interest groups use words that frame their agenda the way they want it. So "Boat People" becomes "Refugees" which becomes "Asylum Seekers", while others may say "Economic Refugees" depending on their frame, and so on.

A handy way to see how word usage changes over time, becoming more or less popular is to use Google's  is to use Ngram Viewer:



SHARE:
© ElasticTruth

This site uses cookies from Google to deliver its services - Click here for information.

Professional Blog Designs by pipdig