27 March 2020

Recycling Needs Recycling

Welcome back. In the 1970s, our group’s research program at Cornell undertook an unplanned, multifaceted effort to address landfills, aka solid waste disposal sites.
Example landfill papers and EPA report by Cornell.
We first assisted a county in selecting sites for a new landfill, our contribution being based largely on a professor’s unparalleled expertise in airphoto interpretation of landforms. We then assisted New York and EPA in developing airborne remote sensing methods to detect seepage (leachate) from active landfills for sampling and testing.

And then came Love Canal. Beginning with the state’s telephoned request for consultation before the toxic industrial waste site made the national news, we acquired and analyzed background information and some 12 dates of aerial photographs to develop a detailed historical record of the abandoned canal, drainage, residences and other structures.

Stereograms of 3 dates of aerial photographs of Love Canal landfill: 1938 (top), 1951 (middle) and 1966 (bottom); note water-filled trench (“T”) in 1938 is partially filled in 1951, and the area is fully developed in 1966. North is to the left.
While we were focused on containing, controlling and remediating landfills, the country began Earth Day celebrations, Congress passed the Resource Conservation and Recovery Act to manage municipal and industrial waste, and the push to “Reduce, Reuse, Recycle” was taking hold.

Recycling Problems

Toter 48-gallon recycling
container (photo from
Amazon); see P.S. for
story of recycle symbol).
The idea of recycling caught on big time with Americans, but the practice had failings, as a number of articles have pointed out. (Columbia University’s Earth Institute offers an excellent summary.)

To start, we haven’t been very good about recycling. Many items collected for recycling are not recyclable. Many recyclables become contaminated by being placed in the wrong recycling bin or in the same bin for single-stream recycling or by adding items that haven’t been cleaned.

And we’re still a throwaway culture. EPA reports that only about 35% of the 268 million tons of municipal solid waste we generated in 2017 was recycled or composted--66% of discarded paper and cardboard, 27% of glass, 15% of textiles and 8% of plastics.

Management of U.S. municipal solid waste, 2017
(from www.epa.gov/facts-and-figures-about-materials-waste-and-recycling).
Recycling of plastics is a special problem. Plastics are often contaminated, and consumer-goods companies are averse to purchasing recycled plastic that’s not as pure as new plastic. Moreover, though we may toss a wide range of plastics into our recycling bins, Greenpeace reports that only some PET (#1) and HDPE (#2) plastic bottles and jugs can be legitimately labeled as recyclable in the U.S. today. The rest is not recycled.

The China Connection
The problem that brought matters to a head is that we got into the habit of shipping our recycling to China--16 million tons of plastic, paper and metals in 2016--where they were used in manufacturing or disposed of. 

For decades, China was a major importer of recycled materials from the U.S. and other countries (from www.latimes.com/opinion/story/2020-01-27/us-shouldnt-let-china-win-the-plastic-trash-war).
In 2018, China cut global imports of plastics to near zero and of mixed paper by a third. Recycled aluminum and glass were less affected. When other countries stopped filling in, we found our own recycling infrastructure was ill prepared.

Lacking a federal recycling program, our processing facilities and municipalities have had to pay more to recycle or cut back and discard the waste.

Wrap Up
There is reason for cheer. The global recycling market for paper, cardboard and plastic is expected to grow. Companies are stepping up efforts to improve and make greater use of recycled plastic. And earlier this month, the U.S. Department of Energy announced up to $25 million in funding for plastics recycling research and development. The funding opportunity is part of the Department’s Plastics Innovation Challenge, a program to accelerate innovations in plastics recycling technologies.

To further support recycling, Columbia’s Earth Institute recommends improving the technology for sorting and recovering materials, incorporating more recycled material into products, getting these products into the marketplace and creating demand for them.

Now, we just have to transform our throwaway culture to one that reduces, reuses and recycles. Thanks for stopping by.

P.S.
Love Canal:
archive.epa.gov/epa/aboutepa/love-canal-tragedy.html
en.wikipedia.org/wiki/Love_Canal
Earth Day: www.earthday.org/history/
Resource Conservation and Recovery Act:
www.epa.gov/rcra
www.govinfo.gov/content/pkg/STATUTE-90/pdf/STATUTE-90-Pg2795.pdf
Reduce, Reuse, Recycle:
pantheonchemical.com/reduce-reuse-recycle/
www.epa.gov/recycle
americanhistory.si.edu/collections/search/object/nmah_1284430
Recycling symbol designer: logoblink.com/img/2008/03/recycling_symbol_garyanderson.pdf
Recycling problems:
blogs.ei.columbia.edu/2020/03/13/fix-recycling-america/
www.greenpeace.org/usa/wp-content/uploads/2020/02/Greenpeace-Report-Circular-Claims-Fall-Flat.pdf
www.epa.gov/facts-and-figures-about-materials-waste-and-recycling
theweek.com/articles/819488/america-recycling-problem-heres-how-solve
www.usatoday.com/story/news/politics/2017/04/20/weak-markets-make-consumers-wishful-recycling-big-problem/100654976/
www.nytimes.com/2019/03/16/business/local-recycling-costs.html
www.bloomberg.com/quicktake/recycling-crisis
China’s ban on importing recycled plastics:
e360.yale.edu/features/piling-up-how-chinas-ban-on-importing-waste-has-stalled-global-recycling
www.theatlantic.com/technology/archive/2019/03/china-has-stopped-accepting-our-trash/584131/
DOE funding for plastics recycling research: www.energy.gov/articles/department-energy-announces-25-million-plastics-recycling-rd-launches-upcycling-consortium

20 March 2020

Yielding to Pedestrians

Welcome back. In a blog post here a few months ago, we learned that physicians of certain specialties were more likely to be ticketed for extreme speeding (20 mph over limit) and that luxury car ownership by speeding-ticketed physicians also differed by specialty (Which Physicians Speed?). At any rate, that's what goes on in Florida.

A midblock crosswalk
(from www.pe.com/2019/11/10/).
Today, we visit Las Vegas, where the urban design is characteristic of sprawl, including automobile-dominated development with long blocks and frequent midblock crosswalks.

Here, our focus is not speeding physicians but whether drivers yield to pedestrians at midblock crosswalks, as required by Nevada law. (Preview: Most drivers don’t.)

This photo of a Los Angeles crosswalk illustrates how drivers don’t always yield to pedestrians and how drivers might get impatient when pedestrians just keep coming (from www.losangeleswalks.org/los_angeles_pedestrian_bill_of_rights_1987).
Midblock Crosswalk Study
University of Nevada, Las Vegas, researchers conducted pedestrian crossing experiments at two midblock crosswalks. The two locations were similar in street design (zebra-striped, non-signaled, 35 mph speed limit, 4 vehicle lanes with a center turn lane) and area income characteristics.

The experiments took place 10 a.m. to noon on a Saturday and Sunday. Four participants acted as pedestrians attempting to cross the street at the crosswalks (two females, one white, one black, and two males, one white, one black).

The participants wore red T-shirts, were briefed on safety protocol and instructed on approaching the crosswalk and crossing the roadway. They were to cross only when no other pedestrians were present, step at least one foot off of the curb to indicate a clear intent to cross and attempt to make eye contact with the driver. The crossings were video recorded for analysis.

Driver Yielding Analysis
Each participant made 30 crossings at each midblock crosswalk; however, technical problems reduced the number of video-recorded crossing attempts to the following: Street 1 – black female, 22, white female, 29, black male, 0, white male, 29; Street 2 – black female, 30, white female, 23, black male, 27, white male, 30.

The researchers analyzed the video data to assess if vehicles yielded for the pedestrians and to estimate the cost of the vehicles based on Kelly Blue Book pricing categories (vehicle make, model, year and condition).

They applied a generalized linear mixed model to evaluate if the average vehicle cost (in thousands), pedestrian gender, pedestrian race and street location were associated with whether or not drivers yielded.

Wrap Up

Of 461 vehicles observed, only 129 (28%) yielded to pedestrians. Cars yielded more frequently for females than males (31% vs 24%) and for whites than blacks (31% vs 25%). Vehicle cost was a significant predictor. The odds that a driver would yield dropped about 3% for every $1000 increase in vehicle cost.

Discussing the results, the researchers state that, without interviewing the drivers, it’s impossible to understand the reason for not yielding. I agree, yet, having stood by the roadside doing traffic counts when I was an undergrad, I thought they could have at least tried to tally the gender, race and general age of the drivers to gauge any effect on yielding.

Nevertheless, the results are in line with earlier studies and point to the need for improved education and engineering and increased enforcement. This is likely to be especially true for midblock crosswalks.

Although this midblock crosswalk lacks a signal, the signs, advance-yield road markings and refuge island are significant safety improvements (from pedbikesafe.org/PEDSAFE/countermeasures_detail.cfm?CM_NUM=13).
Thanks for stopping by. Drive safely.

P.S.
Study of yielding at midblock crosswalks in Jour. of Transport & Health:
www.sciencedirect.com/science/article/pii/S2214140520300359#!
Article on study on EurekAlert! website: www.eurekalert.org/pub_releases/2020-02/uonl-doe022620.php

13 March 2020

Questioning Forensic Psychology

Welcome back. Nearly two years ago, on another website, I described two problems confronting forensic science--the application of scientific principles and techniques to criminal justice (reposted here as Forensic Science Gap).

One problem is that forensic science is, well, short on science. The second problem was that the Trump administration had just terminated an Obama-era initiated commission established to research and make recommendations for improving forensic professionals and laboratories.

Forensic science can be short on science (graphic from www.knowablemagazine.org/article/society/2018/when-courtroom-science-goes-wrong-and-how-stats-can-fix-it).
I’ll pass on revisiting the second problem, but a recent study stirred thoughts about the first problem.

My earlier post addressed forensic methods, such as bite mark analysis, which evolved outside of traditional science before Supreme Court rulings led to the requirement that scientific evidence must be both reliable and relevant.

The recent thought-stirring study concerns forensic psychology, where psychological science or professional practice is applied to help resolve legal, contractual or administrative matters. It appears that forensic psychology is also short on science.

Forensic psychologist contribution in legal matters
(from Forensic Psychology Essentials webinar
www.youtube.com/watch?v=6efx-01bILs).
Forensic Psychology Study
A team of researchers affiliated with Arizona State, Vanderbilt, California and Nebraska-Lincoln universities conducted a two-part investigation.

They first reviewed the scientific basis of 364 psychological assessment tools identified in 22 surveys of mental health practitioners as having been used or acceptable for use in forensic settings.

The second part of the investigation analyzed legal challenges to admitting evidence provided by these tools.

Psychological Assessment Tools
The 364 psychological assessment tools reviewed by the researchers included:
- aptitude, achievement, personality, psychological and diagnostic tests;
- measures designed for both adults and youth; and
- tools that can be used to address referral questions (e.g., competence or fitness to stand trial, violence or sexual offender risk assessment, mental state at time of offense, aid in sentencing, disability, child custody or protection, civil commitment, civil tort, guardianship, competency to consent to treatment, juvenile transfer to adult court, fitness for duty and capacity to waive Miranda rights).

Forensic psychological assessment testing
(photo from www.sacap.edu.za/blog/psychology/types-of-psychology/).
They could not find manuals for about a quarter of the tools and postulated that about 10% have no manual.

More significant, about 37% of the 364 tools lacked any authoritative reviews in recognized review sources (e.g., Mental Measurements Yearbook). Of the tools reviewed, only about 40% received generally favorable reviews. Reviews of 37% of the tools were mixed, and reviews of 23% were generally unfavorable. Some tools had apparently been published without scientific peer review or scientifically sound testing.

Court Acceptance of Psychological Assessment Tools
To focus the second part of the investigation, the researchers selected 30 of the 364 tools as exemplars. This subset of tools represented a wide range of legal issues and forensic referral questions as well as a variety of general acceptance and quality.

The researchers searched three example years, 2016, 2017 and 2018, and found 372 state and federal cases that had used one of the 30 tools. The tool’s admissibility or the admissibility of testimony relying on the tool was challenged in only 19 of the 372 cases, and only 6 challenges succeeded.

The admissibility challenges in the 19 cases involved 9 of the 30 psychological assessment tools. Of the 9 tools, 5 had favorable reviews, 2 had mixed reviews, and 2 lacked any review but were generally accepted.

Wrap Up
The study found that many of the assessment tools used by psychologists and admitted into legal contexts have weak or unknown scientific foundations. Attorneys rarely challenge the expert evidence, and judges tend not to subject psychological assessment evidence to the scrutiny required by the law.

In short, evidentiary challenges to psychological tools are rare, and challenges to scientifically suspect tools are even rarer or nonexistent. The researchers hope the study will help bring about change.

Thanks for stopping by.

P.S.
Study of psychological assessments in legal contexts in Psychological Science in the Public Interest journal: journals.sagepub.com/stoken/default+domain/10.1177%2F1529100619888860+-+FREE/pdf
Articles on study on Assoc. for Psychological Science and EurekAlert! websites:
www.psychologicalscience.org/publications/psychological-assessment-in-legal-contexts-are-courts-keeping-junk-science-out-of-the-courtroom.html
www.eurekalert.org/pub_releases/2020-02/afps-tvi021420.php

06 March 2020

Predicting Dementia from Medical Records

Welcome back. Other than occasionally forgetting a name or the right word, I guess I’m doing OK for my age. Alzheimer's disease is the most common cause of dementia among older adults. At least half of those living with the disease and related dementias never receive a diagnosis. Many more live with symptoms for years before being diagnosed.
Neuropsychological testing for dementia or Alzheimer’s disease (from dailycaring.com/diagnosing-alzheimers-or-dementia-neuropsychological-testing/).
As yet, there is no cure, but early identification can provide the opportunity for screening and delaying the disease onset. Although current tests to screen for dementia risk are invasive, time-consuming and expensive, there may be a breakthrough.
Brain imaging, such as by CT, MRI and PET, for studying structural and biochemical changes associated with Alzheimer's disease and related dementias (photo of amyloid PET scan from www.medicalnewstoday.com/articles/324877).
Two recent studies were able to predict dementia without specialized testing. The level of accuracy, especially for predicting dementia within one year of its onset, was suitable for automated pre-screening.

Machine Learning of Electronic Medical Records
The two studies were conducted by largely the same team of researchers affiliated with Indiana University-Purdue University at Indianapolis, Regenstrief Institute, Merck & Co., Albert Einstein College of Medicine and Indiana University School of Medicine. Collaborators from Georgia State University and Solid Research Group LLC participated in one of the studies.

Both studies achieved early identification of Alzheimer's disease and related dementia by applying machine learning, a subset of artificial intelligence, to routine medical care data widely available through electronic medical records.

Electronic medical records offer a rich source of data for machine-learning analysis (photo from nursingeducation.lww.com/blog.entry.html/2017/02/14/virtual_simulationa-lMOf.html).
The analyzed data were from the records of patients with dementia (cases) and patients without dementia (controls) of comparable age, race and sex. Most patients were from multiple institutions, and all of the data were from the Indiana Network for Patient Care, a regional health information exchange.

Modeling Approaches
The studies took different approaches to model development.

To predict dementia one year and three years prior to disease onset, one study relied on training and testing with data sets for time periods covering 1 to 10 years and 3 to 10 years. Training for one-year prediction was based on 1,728 cases and controls, while testing used 431 cases and over 9,800 controls. About half the number of cases and controls were analyzed for the three-year prediction.

The second study used area under the receiver operating characteristics curve to determine the best models for 1 to 10 years, 3 to 10 years, and 5 to 10 years. The derivation sample consisted of 10,504 cases and 39,510 controls; the validation sample included 4,500 cases and 16,952 controls.

Structured and Unstructured Data
The data sets used by each study included structured data--drug prescriptions and diagnoses--as well as unstructured data--medical notes on visits, progress and medications. The medical notes, a sequence of records identifying the patient and date, followed by a list of reports, each in free text, were the best source of predictive features.

Each study tested models that isolated and combined the different data sets, and each obtained the highest accuracy with combined structured and unstructured data. The models were found to be unaffected by biases related to race, sex or the source institution.

Wrap Up
The researchers are continuing to explore development of dementia models with machine learning and electronic medical records data.

If model performance is confirmed to be comparable to that based on specialized medical tests, the approach would be accessible to the general patient population at a reduced cost. Minimally, the modeling could be used to pre-screen for targeted medical tests for at-risk patients.

Once again, stay tuned. And again, thanks for stopping by.

P.S.
Key facts about Alzheimer’s disease: www.cdc.gov/dotw/alzheimers/
Biomarkers for dementia detection: www.nia.nih.gov/health/biomarkers-dementia-detection-and-research
Studies of predicting dementia from electronic medical records in Artificial Intelligence in Medicine journal and Jour. of American Geriatrics Society:
www.sciencedirect.com/science/article/pii/S0933365718306481
onlinelibrary.wiley.com/doi/abs/10.1111/jgs.16218
Article on studies on EurekAlert! website: www.eurekalert.org/pub_releases/2020-02/ri-sdn021020.php

28 February 2020

Lie or Be Thought a Liar

Welcome back. Yep, I’m blogging about telling lies again. It’s only been about a month since my last post on lying (Liars, Lies and Lying), but this was too interesting not to share with you (and it’s not about the president).

Most of us care about being or appearing to be honest, right? Yet we sometimes lie. Maybe it’s just to make someone happier--you look fine, when it’s too late to change clothes, the doctor who says there’s always hope, when there is none.

Today’s post reviews a recent study of times we might lie so we don’t appear to be, well, lying. When our results are outstanding, when they seem too good to be true, we may lie about how good they are so we’ll appear honest.

Should you lie if your results are outstanding? (Photo from federalsoup.com/articles/2019/06/26/bill-would-omit-polygraph-requirement-for-certain-cbp-applicants.aspx)
Lying Experiments
Researchers affiliated with The Hebrew University of Jerusalem, University of Chicago and University of California, Los Angeles, conducted a series of experiments on lying to appear honest. The different experiments enlisted from 101 to 225 lawyers and undergraduate students in Israel and the U.S. and from 150 to 532 adult online participants in the U.S. and U.K.

Most of the experiments measured if and how much the participants would lie if some manner of testing gave them perfect rather than average results. Other experiments tested if lying was justified. Would observers judge the participants to be dishonest if their exceptional results were perceived as being too good to be true?

Example Experiments
In one experiment, 115 lawyers were asked to imagine working on a case they had told the client would take between 60 and 90 hours. The lawyers were randomly assigned to one of two outcomes: they ended up working either 60 hours or 90 hours on the case. They then had to report the number of hours they would bill the client.

Most lawyers in the 90-hour condition told the truth; however, 18% underreported, billing fewer than 90 hours. None of the lawyers in the 60-hour condition billed fewer than 60 hours.

In an experiment using computer applications, 225 undergraduates rolled a die 4 times and flipped a coin 8 times, reporting the number of “wins.” The researchers  manipulated the program so that one group got 12 out of 12 wins, while another group got a random number of wins. This experiment was done first with a small monetary award for wins, then repeated with no monetary award.

When no bonus was awarded for wins, 26% of the participants with perfect scores underreported their number of wins. When a bonus was awarded, participants with perfect scores were apparently even more concerned with appearing dishonest; 35% underreported their results.

Wrap Up
Although the desire to appear honest generally leads people to lie less, the study showed that it may cause people to lie more. Those who receive outstanding results in private may report less favorable results in public, so others won’t think they are lying.

Is that concern valid? Do people think those with outstanding results may be lying? Two of the study’s experiments found the more favorable the reported results, the more dishonest the person was judged to be.

Oh, go ahead and be honest. Of course, it won’t hurt to be humble if you score way above average. Thanks for stopping by.

P.S.
Study on lying to appear honest in Jour. of Experimental Psychology: General: www.apa.org/pubs/journals/releases/xge-xge0000737.pdf
Article on study on Eurekalert! website: www.eurekalert.org/pub_releases/2020-01/apa-pml012820.php