You don't need to know this to be able to read the rest of the site so only read this if you're (a) really interested, (b) a bit of a nerd or (c) incredibly bored [delete as apporpriate]

Background to the LIWC

LIWC stands for "Linguistic Inquiry and Word Count." It is a text analysis program developed by James Pennebaker to identify what psychological processes are going on in people's writing. (Pennebaker pronounces LIWC like "Luke" which seems odd. If he were British and spelled "enquiry" with an "e" instead of the US spelling of "i" then pronouncing LEWC like "Luke" would make more sense).

Having found in several experiments that people who wrote about trauma enjoyed better health than people who wrote about neutral or trivial topics (like their shoes or the decor of the room they were sitting in), Pennebaker used the LIWC to see if certain types of trauma-writing were healthier than others. He found that people who began the first writing session with few cognitive words and ended the last session with many more cognitive words were most likely to improve their health.

Cognitive (or "thought") words are words that indicate the presence of psychological processes like "insight" or "causal reasoning". "Insight" is indicated by words like understand, know and think while "causal reasoning" is indicated by words like although and because.

This type of work has gone on to investigate adaptation to bereavement, re-employment after job loss, grades in homesick students and even to compare poems written by poets who went on to commit suicide with those written by poets who did not (I'm also doing some work on songs written by songwriters who went on to commit suicide).

Cognition is only one category of words measured by the LIWC. There are lots of others including positive emotion, negative emotion, communication and other social processes as well as parts of speech like pronouns and verb tense. I've listed some of the more important ones below with some example words (in brackets) that indicate the presence of that process or category. There are 73 categories of words measured altogether and these include over 2000 words or word-stems so this is only a tiny selection.

Positive emotion	accepting, active, admiration, agree, amazed, attachment, attractive, award, beautiful, benefit, calm, certain, cheerful, commitment, confidence, creative, delicious, easy, engaging, flirt, fun, gentle, giving, goodness, grateful, harmony, heaven, impress, inspire, interesting, kiss, lively, loving, nurture, paradise, passion, positive, pretty, prize, profit, relax, romantic, sociable, strong, superior, talented, treasure, value, warmth
Negative emotion	ache, afraid, bitter, boring, careless, cries, defeat, difficult, empty, envious, fail, forbid, greedy, grief, helpless, horrible, indifferent, insult, jealous, kill, liar, lonely, mad, moody, neglectful, nervous, obnoxious, overwhel, pain, panic, rage, rebel, sad, scream, tense, terror, ugly, unfortunate, vain, victim, weak, wicked
Cognition: Causal reasoning	assume, basis, depending, how, infer, justifies, motivation, origin, outcome, purpose, reason, therefore, why
Cognition: Insight	acknowledge, admit, became, believe, clarify, concludes, decide, discover, examine, explain, felt, find, generate, informs, knew, learn, meaning, perceive, question, realize, reconsider, secret, suspect, think, understand, wonder
Social processes	acquaintance, assemble, boyfriend, buddy, celebrate, colleague, crowd, dad, daughter, ex, fiance, friends, game, girlfriend, guest, helped, husband, human, individual, infants, kids, ladies mate, mob, neighbor, organization, partner, people, relatives, sister, sweetheart, uncle, who, wife
Social: Communication	advice, beg, communicate, consult, e-mail, excuse, flatter, give, hello, interact, meet, mention, negotiate, praise, replies, respond, rumor, suggest, telephone
Metaphysical: Religion	angel, bible, catholic, church, fundamentalist, god, holy, jesus, jew, lord, meditate, mercy, moral, pray, priest, religious, sacred, saint, satan, soul, spirit
Metaphysical: Death	ashes, burial, buried, casket, cemetary, coffin, cremation, dead, death, deceased, deterioration, drown, dying, funeral, mortal, murder, suicidal, terminate
Sexual	bi, condom, gay, homosexuality, incest, lesbian, libido, lover, lust, porn, prostitute, queer, sex, stud, virgin

The LIWC counts words and assigns them to these categories, expressing each category as a percentage of the whole text. For example, in a text of 100 words, if there were 20 words like "happy" or "pleased" and 5 words like "sad" or "angry" then the LIWC would indicate that 20% of the text was positive emotion and 5% of the text was negative emotion. In other words, the LIWC represents psychological processes as numbers or statistics.

It really IS that simple! The difficult thing is interpreting what it means.

Before we go any further, there are some myths about statistics that I would like to dispell. People use phrases like "there are lies, damn lies and statistics" or "you can prove anything with statistics". This is CRAP! It's usually people who don't understand statistics that use phrases like these. The truth is you can't prove anything with statistics. Statistics are just a way of describing what is going on in the real world, just as words are another way of describing (qualitatively rather than quantitatively) what is going on in the world - you might as well rewrite these quotes as "there are lies, damn lies and words" or "you can prove anything with words". This is no less a nonsense than claiming this of statistics.

In incompetent or malevolent hands statistics can be uninformative and even misleading. But in competent hands statistics can give clues to real processes that are going on in the world around us.

I emphasise the word "clues". Nothing that I write here is proven, I have merely presented evidence that leads to certain suggestions. Other people may have other ideas to account for the evidence. The statistics are presented as evidence and these are facts - however the interpretation of these may be subjective.

Having said all that, you don't need to have any knowledge of statistics to read the rest of this website. I've tried to present it in an accessible way for any intelligent person to understand (and obviously "intelligent" should include any Bowie fan) regardless of level of familiarity with statistics or love of mathematics. If any of the more nerdy types want more information about the analyses then I have generally included this information in footnotes so that it doesn't interfere with the flow of the text.

Luv on ya

Dr Nick