Monday, August 12, 2019

The Slippery Business of Bias and the Media

Author: Maistriani the Machiavellian

Snow Leopard, endangered species of uncontrolled poaching

Our brother Germaine has asked me to "try" to put out a post on evaluating media reporting with a systematic process. Mind you, I'm wordy, in case none have noticed, and maybe a bit towards the esoteric lean with my writing style and word choices. He promised to beat me with multiple implements of cooked pasta, delete all my posts and ban my account if I didn't bring it down a notch, so I'll try. References will be at the end, for those interested.

BIAS

Seeing as we are talking about media and the language they use, I'll go with a definition of writer's bias:
Bias occurs when a writer displays a partiality for or prejudice against someone, something, or some idea. Sometimes biases are readily identifiable in direct statements. Other times a writer's choice of words, selection of facts or examples, or tone of voice reveals his or her biases.

That's not bad, but I've been really looking into language bias for about the last year solidly and what becomes apparent is that it is quite a bit more involved. We generally tend to look at bias through the socio-cultural prisms of offensive or vilifying statements, propositions or words that directly indicate an attack of a person or social grouping by race, sex, gender identification, religion, etc.

That's the right in your face kind, but what if there is more? Much more, and it is sublime because even though we don't get the conscious "bump - hey that's not right", there is a non-conscious affective outcome that we very often feel as specific emotion towards, but are very unlikely to be able to point to a cause. Here's a short list of what I would term "occult language modifiers":

1. Factive verbs: are verbs that presuppose the truth of their complement clause.

2. Implicative verbs: : implicative verbs imply the truth or untruth of their complement, depending on the polarity of the main predicate. (Polarity here is also called valence, which means positive or negative word, statement, proposition, tone, word choice)

3. Assertive verbs: are verbs whose complement clauses assert a proposition. The truth of the proposition is not presupposed, but its level of certainty depends on the asserting verb.

4. Hedges: used to reduce one’s commitment to the truth of a proposition, evading any bold predictions.

5. Strong subjective intensifiers: are adjectives or adverbs that add (subjective) force to the meaning of a phrase or proposition.

6. Degree Modifiers: are contextual cues (often adverbs such as extremely, or slightly) that modify the intensity or degree of an action, an adjective or another adverb.

Now just using those few markers, we could go through virtually any news media reporting and find bias that we likely never noticed before, probably passed over and weren't actually aware was affective of both our thinking and our mood, especially towards the reported subject or issue.

All of media will be found to be biased, even if that bias goes towards our particular political, social or cultural preferences, it's still bias. Because bias is affective of thinking processes, emotions and moods, it isn't making us better thinkers and problem solvers to allow it to happen.

So what do we do? Change the media? I'll wish you the best of luck with that but don't think you'll manage very well. What we need to do is build better tools for better thinking as the best defense against these problems.

The best part about Biopolitics and Bionews is that the channel has a foundation enforcing that strong ethics are required. Strong ethics is the foundation of creating one of our best tools: the rubric.

RUBRIC

Rubric: A rubric sets out clearly defined criteria and standards for assessing different levels of performance. Rubrics have often been used in education for grading student work, and in recent years have been applied in evaluation to make transparent the process of synthesising evidence into an overall evaluative judgement.

So we start with our ethical decision making and create a rubic built from our ethical values.

The great thing about a rubric that is built from ethical decision making is that it is easily modifiable to accommodate a myriad of different applications.

So, when we are reading a media news story, we can pull up our rubric if we keep it on our PC or MAC, or as I prefer to have it separate because it works better for me, I print it out. Then we can dissect and analyze any media coverage by comparing the statements, facts, commentary to our rubric and ask ourselves the question, "does this meet with my rubric?" Obviously, we can use this across media platforms, media outlets and even for those who are engaged through social media.

METRIC

For some, an extra step may be used, although it isn't absolutely necessary. We can also apply a metric to gauge how well any media coverage has done with regards to facts, accuracy, ethical decisions. The extra component added should be to check for and assess not only the presence of bias, but the actual level of bias.

There is an awesome application called VADER, which means Valence Aware Dictionary and sEntiment Reasoner. Sentiment analysis is becoming more and more important across domains, corporations are even using it to modify their application processes to achieve better outcomes on hiring targets.

Sentiment analysis means not just the measurable positivity or negativity of text but also of biased wording and even that found in text containing slang, emojis, similes, metaphors, etc. It is a whole tool language metric with a proven track record.[*see below] If not your style or too much technology involved, in the references you will find sufficient information for standard metrics used in bias and sentiment analysis.

So this is my toolbox for Defense against the Dark Arts of media bias, dark free speech and failures at missing bias in my own cognitive processing. If Germaine approves this post, hope to see comments and will help with inquiries in any way possible.

REFERENCES
https://www.betterevaluation.org/en/evaluation-options/rubrics
https://blogs.ei.columbia.edu/2013/12/02/sustainability-ethics-and-metrics/ http://oxfordre.com/communication/view/10.1093/acrefore/9780190228613.001.0001/acrefore-9780190228613-e-439
https://www.thinkmind.org/download.php?articleid=huso_2015_2_30_70077
https://research.ku.edu/sites/research.ku.edu/files/docs/EESE_EthicalDecisionmakingFramework.pdf
https://www.victoria.ac.nz/vbs/teaching/aol/rubrics-and-score-sheets/LO-4a-Rubric-for-Ethical-Perspectives.pdf
https://medium.com/analytics-vidhya/simplifying-social-media-sentiment-analysis-using-vader-in-python-f9e6ec6fc52f
[*downside] The only downside is VADER requires operation on a Linux distribution because it uses Python. For anyone interested, feel free to ask, and even if you are a long time Windows or MAC user, there is a Debian distribution called DeepIn that I am confident most would find very comfortable and easy to use. It is also the most beautiful OS I have ever seen.

B&B orig: 1/18/19

No comments:

Post a Comment