UPI Study › Courses › Abnormal Psychology › Free lessons › Lesson 1.4 Free with a free account
Pitfalls in Mental Health Classification
Abnormal Psychology · Chapter 1: Foundations of Abnormal PsychologySection 1.4 · lesson 4 of 9716 min · free with a free account
Free preview
Create a free account to keep watching
It takes a minute and costs nothing. Your free account unlocks all of Chapter 1 and 3 more lessons of Abnormal Psychology.
Lesson complete
Earn the 3 college credits for Abnormal Psychology
Enrolling adds a quiz after each lesson, the graded work and the official transcript. Recommended for credit by NCCRS.
Next lesson: Mind-Body Connections: How Psychology Shapes Physical Conditions →← The DSM Explained: Purpose and StructureNext lesson: Mind-Body Connections: How Psychology Shapes Physical Conditions →
In this lesson
Accuracy, consistency, and social harm - and why the system is kept anyway.
- Define reliability and validity as they apply to diagnosis
- Explain what comorbidity and heterogeneity imply about category boundaries
- Separate public stigma from self-stigma, and name their measured effects
- State the comparative argument for retaining classification
Transcript
Everything said in the video, section by section (2,240 words). Read along, or search it.
Foundations of Abnormal Psychology
Pitfalls in Mental Health Classification
The last section described how the manual is built. This one asks how well it works. The answer is uncomfortable, and worth sitting with. We start with two words that sound similar and are not. Reliability, meaning consistency. And validity, meaning whether the category is real. Then two structural problems that both point at the same weakness. Comorbidity and heterogeneity. Together they suggest the boundaries may be wrong. After that, the social costs. Bias, labeling, and stigma. Then a threshold that trades one error against its opposite. And finally the question you should be asking by then. If all of that is true, why is the system kept at all?
Before You Begin
The Writing Has to Model What It Teaches
One note before we begin, and a second note specific to this section. This material is educational, and not a basis for diagnosing yourself or anyone else. Now the second note. Part of what follows is about the harm that language does. That means the writing here has to model what it teaches. People are described as people who have a condition. Never as instances of it. Stigmatizing framings are described in their effects, rather than reproduced. When you write about this material yourself, the same standard applies. Say what a framing does to people. Do not repeat the framing to show it.
Reliability
Do Two Clinicians Reach the Same Conclusion?
Reliability, applied to diagnosis, is simply consistency. Does the same presentation produce the same diagnosis in different hands? The form that matters most here is interrater reliability. Two independent clinicians assess the same person. Do they reach the same conclusion? Agreement is not measured as a simple percentage, and there is a good reason. Two clinicians who both say absent to almost everyone will agree almost all the time. They have demonstrated nothing, because agreement by chance alone is already high. So field trials report a statistic that corrects for chance. Near zero means they agree no more than they would by guessing. Two honest observations follow, and both cut against reassurance. First, agreement varies a great deal by category, and the average conceals that spread. Second, field trial conditions are more favorable than ordinary practice. Those clinicians are trained for the trial, assessing deliberately, and know they are watched. Reliability in routine settings is generally lower than trial values suggest.
Validity
Necessary, and Nowhere Near Sufficient
Validity is the other question, and it is the harder one. Does the category correspond to something real? Three forms matter here. Construct validity asks about a coherent underlying condition. Evidenced by symptoms that cluster, a characteristic course, and shared mechanisms where possible. Predictive validity asks whether the diagnosis forecasts anything. That form carries the most practical weight. A category that predicts nothing cannot guide a decision. Concurrent validity asks whether the diagnosis agrees with other measures taken at the same time. Now the relationship between reliability and validity, which is asymmetric. Reliability is necessary for validity. A category clinicians cannot apply consistently cannot be measuring anything stable. But reliability is not sufficient, and that is where people go wrong. Clinicians can be trained to apply a rule consistently whether or not the rule carves anything real. The upper left corner is unreachable, because inconsistency blocks any claim to validity. The lower right corner is entirely reachable. Perfect agreement about an empty category. The reforms of nineteen eighty improved reliability substantially. They settled far less about validity than is sometimes assumed.
One Presentation
Two Clinicians, and a Divergence With a Cause
Make that concrete with a composite built for instruction, describing no real person. Samuel is twenty-nine, and is assessed twice, eleven days apart. He describes periods of very low mood, disturbed sleep, and irritability. And in the past year, two stretches of about five days each. In those stretches he slept little, felt unusually capable, and began several projects. Difficulties have been present for roughly three years, and he drinks heavily during the low periods. Clinician one concludes a depressive condition with harmful alcohol use. She interviewed him unstructured, in a single fifty-minute appointment, during a low period. She asked about elevated mood in general terms, and he did not volunteer the five-day stretches. Clinician two concludes a bipolar-spectrum condition. He used a structured interview, and saw Samuel across two appointments. He spoke with Samuel's sister, with consent. And he asked about elevated periods using specific behavioral anchors. Now name the source of the divergence, because that is the skill. This is not a difference of opinion about a threshold, and it is not bias. It is a difference in the information available, produced by a difference in method. Contrast a dispute over whether impairment is marked. No extra information settles that one.
Comorbidity
The Rule Rather Than the Exception
Comorbidity is the presence of two or more diagnosable conditions in the same person at once. Its significance is not that it happens. It is that it is not unusual. In clinical samples, the person with exactly one diagnosis is closer to the exception. Look at a constructed sample of three hundred people, invented for teaching and drawn from no study. One hundred and twenty meet criteria for Condition A. Only thirty-eight of them have that diagnosis alone. Eighty-two carry at least one more. Across the whole sample, one hundred and five people have a single diagnosis. One hundred and ninety-five have more than one. Roughly two people in three. Forty-five people meet criteria, and only nine of them appear alone. A category in which almost nobody appears alone raises a fair question. Is it a separate condition, or a feature of the conditions it accompanies? Treatment research usually recruits people with one condition and excludes the rest. So the evidence base is built on samples systematically unlike the people clinicians see.
Heterogeneity
The Same Label, Different People
Heterogeneity is the variation among people who carry the same diagnosis. It follows directly from how criteria sets are written, and the arithmetic is unavoidable. Suppose an entry requires any five features from a list of nine. Two people can each meet that entry while sharing only one feature between them. In some categories they can share none at all. A criteria set of that shape defines a large family of distinct symptom combinations. Clinically, this means the diagnosis tells you less about the person than the label implies. Which is why competent practice records which criteria were met, not just the name. In research, a trial saying a treatment helps may be describing one combination within the category. Averaging across combinations conceals a large benefit for some, and none for others. One says the categories may be too wide. The other says they may be cut in the wrong places. On comorbidity, be careful. Three explanations are live and the evidence does not choose. The conditions may be distinct, with one genuinely raising the risk of another. The categories may be cutting a single difficulty into pieces. Or different conditions may simply share common causes.
Kinds Or Degrees
Categories Against Dimensions, With the Stakes Attached
Now the categorical and dimensional positions, restated with the evidence attached. The case for categories starts with how decisions actually get made. Treat or do not. Admit or do not. Fund or do not. Categories match those decisions, and are learnable in a way score profiles are not. And the entire existing evidence base is organized by category. Abandoning them discards the accumulated literature rather than reinterpreting it. The case for dimensions starts with the data. Symptom counts distribute smoothly, with no natural break where the thresholds sit. People just below a threshold resemble people just above it, in course and in outcome. Comorbidity is pervasive, and heterogeneity within categories is large. A dimensional scheme would capture subclinical difficulty that categories discard. The dimensional position has the better of the argument about what the data look like. The categorical position has the better of the argument about what a working system must do. Two things remain genuinely unsettled, and no evidence available answers them. Whether any psychological condition has a natural boundary at all. And whether a dimensional system could work in services and courts that require decisions.
Rates By Group
Diagnostic Bias, and Four Competing Explanations
Rates of certain diagnoses differ substantially between demographic groups. Some conditions involving psychosis are diagnosed considerably more often in some minority ethnic groups. While some mood and anxiety conditions are diagnosed less often. The observation itself is robust and repeated. What produces it is contested. Four explanations, and each deserves its strongest form. First, clinician bias. Assessment is a judgment made by a person. Expectations about groups can influence which features are noticed and which threshold is applied. Supporting evidence includes studies where identical case material is rated differently by group. Second, differences in expression and communication. An assessment across a linguistic difference can misread ordinary expression as symptom. Third, differences in real exposure. Groups differ in exposure to social adversity, discrimination, migration, and poverty. On that account, part of the difference in rates is a real difference in incidence. Where a group more often arrives through crisis or involuntary routes rather than primary care. The presentation seen at first assessment differs systematically, and that alone shifts what is recorded. Two things can be said with confidence. The mechanisms are not mutually exclusive. And their relative weight is genuinely unresolved. This course will not settle it. Assessing across a group difference, treat your own confidence as less informative than usual.
Labeling
A Famous Study That Does Not Hold Up
A label changes how a person is perceived, and how they are treated. That effect is distinct from whatever the diagnosis itself describes. In a study published in nineteen seventy-three, David Rosenhan reported the following. A small number of people without psychological conditions presented at psychiatric hospitals. Each reported a single brief unusual experience, and otherwise behaved normally. According to the report they were admitted, mostly with a diagnosis of a psychotic condition. And their ordinary behavior, including note-taking, was recorded in terms fitting the diagnosis. They were discharged with the condition described as in remission, rather than absent. Now the criticisms, which are substantial and must be given with the study. Presenting a symptom that was not real means the clinicians were deceived about their only evidence. Admitting someone who reports an unusual experience is arguably correct practice, not error. The study had no control condition. Much of the reported detail rests on the accounts of the participants themselves. And the discharge language is in fact the standard way of recording that someone is currently well. Then in twenty nineteen, the journalist Susannah Cahalan published investigative work. It raised serious doubts about the underlying records, including whether all participants existed. So where does that leave us? Separate the proposition from the study. The study is not sound evidence for its headline claim, and should not be cited as though it were. The narrower claim, that a label shapes how later behavior is read, survives on other evidence. That separation is the lesson. Look for better evidence, rather than discarding the question.
Stigma
What Others Do, and What a Person Absorbs
Stigma is the set of negative attitudes, beliefs, and behaviors directed at people. Two forms are distinguished, and the distinction matters because the responses differ. Public stigma is what other people believe and do. The stereotypes held about a group, the emotional reactions those produce, and the behavior that follows. Reduced willingness to work alongside, to rent to, or to live near people with certain diagnoses. And a documented disadvantage in employment for applicants whose history is known. Self-stigma is what happens when a person absorbs those beliefs and applies them to themselves. Its measured effects include reduced self-worth and reduced confidence in one's own capacity. And this one matters most. Reduced engagement with treatment. Because seeking help means accepting the identity the person is trying to avoid. Three consequences are consistently documented across both forms. People delay seeking help, often for years after difficulties begin. People conceal a diagnosis from employers, colleagues, and sometimes from clinicians. And people withdraw from social contact in anticipation of rejection. Which removes exactly the support that protects against deterioration. It is produced by the expectation of rejection, whether or not rejection occurs. Which is why measuring stigma by counting hostile acts understates it.
One Threshold
Opposite Failures, and Why the System Is Kept
Two opposite errors remain, and they are governed by one dial. Overdiagnosis is assigning a diagnosis where the difficulty does not warrant it. Underdiagnosis is failing to identify a condition that is present. Here is the part people miss. Reducing either one increases the other. The costs of overdiagnosis include unnecessary treatment, and its side effects. Stigma and self-stigma. Diversion of limited resources from people with greater need. And the reframing of an ordinary difficulty as a medical one, obscuring what produced it. The costs of underdiagnosis include untreated suffering, and avoidable deterioration. Exclusion from services that require a diagnosis to access. Those costs are not symmetrical, they differ by condition, and they fall on different people. Which makes where to set a threshold a value judgment as much as an empirical one. So read the whole section together, and a reasonable reader might want to abandon the system. The argument against that is a counterfactual. Consider what happens with no shared system. Clinicians have no compressed way to communicate a presentation. Treatment research has nothing to be about, because a trial needs a defined population. Services cannot be planned, because there is no way to count who needs what. And access to funded care would rest on individual judgment with no external reference. People would still be classified anyway. Informally, by non-experts, in harsher language. So the honest conclusion is not that the system is good. It is that this is currently the best available instrument for jobs that have to be done.
About Abnormal Psychology at UPI Study
Abnormal Psychology (Psychology 180) is a self-paced online course of 97 video lessons in 25 chapters, from the main theoretical perspectives and clinical assessment to anxiety, mood, psychotic and personality disorders, treatment and the law. It carries a NCCRS credit recommendation for 3 semester credits.
Every lesson transcript is free on pages like this one, and the videos are free with a free UPI Study account. Enrolling adds a practice quiz after each lesson, the graded work and final exam that earn the credit, lifetime access, and an official transcript on completion. UPI Study has helped over 130,000 students with college credits. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Enroll now → Course details Download the syllabus All Psychology and social science courses Every course with PRO →
Why take it here
Why students take Abnormal Psychology with UPI Study
Where the credit goes
Universities that work with NCCRS
Abnormal Psychology carries a NCCRS credit recommendation. These universities work with NCCRS; some are UPI Study's signed partners. ACE and NCCRS have over 2,100 cooperating colleges that will consider accepting credits; the final decision is with the universities.
Thomas Edison State UniversitySigned partner
University of the PeopleSigned partner
Charter Oak State CollegeSigned partner
Strayer UniversitySigned partner
Columbia Southern UniversitySigned partner
Indiana TechSigned partner
Penn StateWorks with NCCRS
Southern New Hampshire UniversityWorks with NCCRS
SUNYWorks with NCCRS
+many more cooperating universitiesFind my college →Find my college → Not listed? Ask Athena in the corner of this page, or book a 15-minute call.
FAQ
Questions students ask us
About Pitfalls in Mental Health Classification
What is the difference between reliability and validity in psychiatric diagnosis?
What does comorbidity mean in mental health?
What is the difference between public stigma and self-stigma?
Is the Rosenhan study still considered reliable evidence?
About the Abnormal Psychology course
How many college credits is Abnormal Psychology?
How much does Abnormal Psychology cost and what is included?
Which universities accept UPI Study credits?
How long does Abnormal Psychology take?
How are the exams done?
How do my credits reach my college?
Ready for the full course? Earn 3 credits in Abnormal Psychology.
97 video lessons, a quiz after each one, and the graded work that earns the credit. Recommended for credit by NCCRS.
3 credits · NCCRSEnroll →