Topic 3: Validity {by 9/23}

Based on the text readings and lecture recording due this week consider the following two discussion points: (1) Discuss your understanding of criterion-related validity (also known as: Prediction or Instrument-Criterion).  In you discussion, include why this particular type of validity is common/important for mental health assessments.  (2) First, discuss the difference between convergent evidence validity and discriminant evidence validity.  Second, provide an example of a hypothetical or real assessment where these two types of validity would be important – very common for many mental health assessments (hint: listen to my example in the lecture recording).

 

Your original post should be posted by 9/23.  Post your two replies no later than 9/25.  *Please remember to click the “reply” button when posting a reply.  This makes it easier for the reader to follow the blog postings.

63 Comments (+add yours?)

  1. Kristin Blair
    Sep 20, 2021 @ 09:49:32

    Criterion-related validity is an estimate of how an instrument agrees with a gold standard, or basically, how well an assessment predicts an outcome for another assessment. An example of this would be, grad students taking the GRE (that WE didn’t have to take so, that’s cool). The score obtained (received now) on the GRE is an attempt to see how well the student will perform in grad school (criterion-cannot be assessed until later). This type of validity is important for mental health assessments because, one of the most considerable concerns is how well it’s actually measuring what we want it to measure. We are also concerned about people’s mental states now, as well as the future. This can show if a person in “on track” to a certain behavior. Dr. V mentions it in terms of assessing for depression in May and then again in October. Reevaluating in October and seeing the scores, can help us validate this behavior when the scores from back in May are examined. One could say, “a certain percentage of people who score above a certain score, whatever that cut off is, are shown to have a higher chance of following through on suicidal thoughts.” These types of patterns are crucial to know in the mental health field.

    Convergent evidence validity and discriminant evidence reliability are both considered parts of construct validity. Convergent evidence validity tests that things that should be related, are in fact related. Discriminant evidence validity tests that things that should not be related, are not related. Even though they are terms for describing opposite findings, neither one alone is sufficient to show construct validity. This is very helpful because there tends to be many symptoms or traits that we try to assess in mental health that overlap with others. Using this distinction can help isolate them and help providers offer more accurate diagnoses and treatment to clients. Dr. V gives an example of this by saying that if he wanted to develop a new instrument for assessing depression, he would also give the individual a more established instrument like the BDI II or the Hamilton. In doing so, he would want a high validity coefficient, if doing discriminant evidence, he would want a low validity coefficient. This will show whether your new instrument is measuring what it’s supposed to compared to other similar ones, or it’s measuring what it’s supposed to compared to others that are supposed to be different.

    Reply

    • Will Roche
      Sep 20, 2021 @ 16:13:58

      Kristin,

      I really like how you presented Dr. V’s example of convergent and discriminant evidence with low and high validity coefficients. In undergrad, I was conducting an experiment where I had to create my own instrument to measure a certain construct
      (claimed self-handicapping). I needed to make sure my own instrument had similar results to a different claimed self-handicapping measure to confirm a high validity coefficient. Another form of self-handicapping is behavioral self-handicapping. In theory, I could have (and probably should have) tested my claimed self-handicapping scale against a behavioral self-handicapping scale. By doing so, I would have been able to determine if there is any discriminant evidence. Hypothetically, if my instrument had a low validity coefficient with a behavioral self-handicapping scale, that would have also strengthened my instrument that I created.

      Reply

    • Victoria Cestodio
      Sep 21, 2021 @ 14:41:45

      Kristin,
      I really loved your example of criterion-related validity when it comes to the GRE. Most schools think that is a great predictor on how well the student will perform in graduate school.
      You make a very good point in the second part of your discussion about discriminant validity. It is very true that a lot of symptoms overlap within mental health. Using this does make us isolate them more and can highly benefit us as mental health professionals. This then provides us with the right diagnosis and treatment to do from there.
      Great response!!
      Victoria

      Reply

    • Olgena
      Sep 25, 2021 @ 21:52:38

      Hello Kristin,

      Your post was so well written and explanatory! It reminded me my capstone class in undergraduate where I had to choose the proper instrument to measure the variables of my research topic. I agree with you where you mention Dr V example it was very helpful to understand the difference between convergence evidence and discriminate evidence, summing that both this methods are different in terms of measuring items but similar how they are structured. Furthermore, their overcome and score results is truly important for a clinician to evaluate the validity and accuracy of these instruments.

      Reply

    • Pilar
      Sep 25, 2021 @ 23:20:33

      Hello Kristin,
      I really liked your definition of criterion related validity. I had much better time understanding your definition versus the textbooks definition. I also liked your use of example.
      Your example of discriminant versus convergent validity using the Hamilton and the BDI was an excellent display long the differences and I also like how you included the use of the terms low validity versus high.
      Thank you for this post!

      Reply

  2. Emily Barefield
    Sep 20, 2021 @ 11:42:24

    Criterion-related validity is simply how well an instrument predicts what it is supposed to predict. Another way to think about criterion-related validity is how well are the scores of the instrument related to the outcome it intended to measure. An example discussed in the video used an instrument intended to measure suicidality. For a suicidality instrument to have criterion-related validity, the scores obtained on the instrument must highly correlate with the defined outcome which would likely be suicidal actions (or potentially thoughts). If an individual scored high on the suicidality measure, we would expect them to have attempted suicide or perhaps report frequent suicidal thoughts for the measure to have criterion-related validity. Concurrent validity is a type of criterion-related validity in which there is no time lag between when the instrument is given and when the outcome is measured. For example, if an individual was given the hypothetical suicidality measure discussed above, and then immediately was asked about their suicidal thoughts we would expect someone who scored high on the measure to report a high level of suicidal thoughts if the measure has high concurrent validity. Predictive validity is a type of criterion-related validity in which there is a lag between when the instrument is given and when the outcome is measured. Defining what is meant by a lag is somewhat subjective and dependent on the test. For example, if we gave the suicidality measure to someone and then six months later, they had attempted suicide, we would say that the instrument has high predictive validity.

    Criterion-related validity is important in mental health assessments because it can, and should, inform treatment. For example, if an individual does score high on a suicidality instrument that has been shown to have a high-criterion validity, that should certainly inform treatment. We would want to implement an intervention that helps to reduce suicidality. Or perhaps, we have an instrument with high criterion-related validity that predicts panic attacks in individuals with panic disorder. If they score high on this instrument, we will want to give them the tools to cope with a panic attack as well as tools that might help them avoid panic attacks in the future.

    Convergent validity means that an instrument intended to measure a certain variable correlates with other variables to which it should theoretically be positively related. For example, if I were to create a new scale to measure anxiety, I would want it to be highly correlated with a commonly used anxiety measure, such as the GAS-7. Divergent validity means that an instrument is not correlated with variables from which it should differ. For example, I would want my anxiety scale to not be correlated with the Beck Depression Inventory.

    Reply

    • Will Roche
      Sep 20, 2021 @ 16:25:51

      Emily,

      You did a great job in detailing the specifics of criterion-related validity as well as concurrent and predictive validity. The suicidality example really helped me understand the general concept as well as concurrent and predictive validity aspect as well. A mirroring example of this would be the GRE. The GRE is supposed to measure the relationship between success on the GRE exam, and strength of graduate schools that the student is accepted into. Hypothetically if a student took the GRE and did extremely well and all 10 schools he applied to immediately assessed his results and accepted him, this would show concurrent validity. In the usual scenario, if the student did very well on the GRE and then applied to those same 10 schools 6 months later, if all those schools accepted him, this would be evidence of predictive validity.

      Reply

    • Vanessa Nichols
      Sep 21, 2021 @ 11:25:06

      Hi Emily,
      I think that your posts are always a great read and informative. I think you did an excellent job explaining Criterion-related validity; A suicidality measure was a perfect example of why this validity is so essential. Suppose the measure does not measure suicidality. This could be so dangerous. Best-case scenario, you talk about suicide and suicidal thoughts with someone who isn’t suicidal, but the worst case is that there is potential for people who are suicidal to slip through the crack.
      I completely agree with you that criterion-related validity is important because it should inform treatment, and I also think it tells the clinician of the severity. The predictive nature allows the clinician to know if action needs to occur right away.
      I think you did a great job describing and differentiating convergent evidence and discriminant evidence.
      Thank you!

      Reply

    • Mary Altomare
      Sep 21, 2021 @ 12:00:27

      Hi Emily,

      I thought your post this week was thorough and well written. I enjoyed your interpretation of Dr.V example of criterion-related validity. As you shared, if an individual does score high on a suicidality instrument that is proven to have a high-criterion validity, the clinician should utilize this when creating a treatment plan for this particular client struggling with suicidal ideation/thoughts. I thought your description of convergent evidence and discriminate evidence was clear and concise. It is important for a new instrument that is created to measure depression has a high validity coefficient to a well established depression inventory, in order to verify that it is measuring the correct variable, and is a reliable instrument to utilize. Whereas, a new depression inventory should have a low validity coefficient with an anxiety inventory instrument, in order to prove that the two instruments are indeed measuring two different variables. Therefore, because criterion-related validity is commonly utilized in the mental health field it is imperative to understand the difference between convergent evidence and discriminate evidence.

      Reply

    • Olgena
      Sep 25, 2021 @ 22:09:41

      Emily,
      Your post was one of my favorites to read this week because made the readings and lecture clearer to understand. I would like to highlight two points that I think are important to understand when it comes to the results and overcome one’s must expect regarding convergent instrument and discriminant instrument. Choosing two instruments that are measuring the same items as in convergent instruments case one must expect that the scores between the two instruments are high. In contrast, using diverse instruments method one’s should expect that the score between the two instruments are low so they are valid and reliable.
      Thank you

      Reply

    • Kelsey McGinness
      Sep 26, 2021 @ 09:11:15

      Emily, I like the examples you provided for criterion-related validity. You expressed your reasoning clearly and complete and helped me further understand criterion-related validity. You are right, the assessments we use, we do not only use them to measure various areas in mental health but also the then further provide the best care and treatment we can for those patients.

      Reply

  3. Will Roche
    Sep 20, 2021 @ 16:02:19

    Criterion-related validity or Prediction/Instrument criterion is used to make a prediction of how certain a specific outcome of a behavior or a construct will be. In the textbook, they use the example of the GRE or the ASVAB. The GRE is used to measure/predict how successful a student would be in graduate school. (I personally took the GRE two years ago and I’m hoping my results are not indicative of my graduate success). Ideally, specific instruments can be used to predict specific behaviors, constructs, successes, etc. In the mental health field, criterion-related validity is so vital because it aims to predict the progression of certain emotions or behaviors of a client, such as depression or anxiety. There are two forms of criterion-related validity: concurrent and predictive validity. If a therapist was searching for concurrent validity for depression, they would use an instrument such as the BDI, and then immediately ask the clients about their lack of interests, feeling blue, etc. If their BDI score was high, and they also noted lack of interest, motivation and frequent sadness, this would indicate strong concurrent validity. If they took the BDI, and returned in six months presenting strong signs of depression, this would indicate strong predictive validity in the instrument being used. This shows why criterion-related validity is so vital in the mental health field.

    When looking at evidence based on variables that may be related to one another, it is important to know the differences between convergent and discriminant evidence. Convergent evidence means an instrument is able to measure a construct similarly to another instrument measuring the same thing. An example would be that the Hamilton Depression Rating Scale and the Beck Depression Inventory would yield similar results if given to the same client. Contrastingly, discriminant evidence would want to predict the opposite, that two tests measuring two different constructs would not yield the same results, (BDI and GAD-7). These types of evidence are important for clients when measuring to determine the relationship between constructs, or lack thereof.

    Reply

    • Mary Altomare
      Sep 21, 2021 @ 12:14:44

      Hi Will,
      I liked your example of the GRE to explain criterion-related validity. I personally did not take GRE, but if I had to guess my score would have been low, just because I typically do not perform well on standardized test. However, I still think GRE is beneficial in predicting whether or not an individual will be successful in graduate school. I agree with you completely, that in the mental health field, criterion-related validity is imperative because it does predict an individual current functioning and future functioning, such as anxiety and depression. This is so important because it helps provide clinicians with important information of their clients current functioning and the best way to treat them!

      Reply

    • Victoria Cestodio
      Sep 21, 2021 @ 14:51:38

      Will,
      I like how you touched on the two forms of criterion-related validity. You used great examples to decipher the two. As I read and listened to the lecture the two were a little confusing to me but how you explained them helped me! I think both types are really beneficial even though they go about things in different ways. I think for some clients concurrent validity would help them best and some predictive validity would help them best. I like that criterion- related validity has these two components so as a therapist we can decide which one would be best for the client.

      Reply

    • Jordan Ellis
      Sep 26, 2021 @ 06:59:38

      Hi Will,
      What a well-written post! I really enjoyed reading it. Thank you for providing examples of concurrent and predictive validity. They helped me to understand the two different concepts. I think the BDI is an assessment that is great for measuring current attitudes. These results can also help us predict a client’s future attitudes and symptoms of depression. Thank you for highlighting the differences between convergent and discriminant evidence, as well. You provided helpful examples of both. It is incredibly important that assessments have these types of validity, and don’t just have face validity.

      Reply

  4. Mary Altomare
    Sep 21, 2021 @ 10:59:58

    Criterion-related validity also known as Prediction or Instrument-Criterion, is essentially how well an instrument predicts what it is assessing for. In other words, the overall instrument is reliable in predicting one’s behavior in the present moment and in the future. Dr. V gives an example of an instrument used to measure suicidality. The instrument is used to measure/predict if an individual is feelings suicidal or engaging in suicidal behaviors based on the score they receive. Therefore, if an individual scored high on the suicidality measurement, mental health counselor could predict that the individual may have either attempted suicide, has a plan or is experiencing suicidal thoughts. This particular type of validity is common/imperative for mental health assessments because it provides mental health counselors an understanding of their client’s mental health status (areas of particular concern. For instance, if an individual obtained a high score on an instrument measuring social anxiety, the mental health counselor will be able to create a treatment plan that aims to reduce the individual’s social anxiety through various therapeutic activities and coping strategies. Furthermore, criterion-related validity is imperative because it does assist mental health counselor in the way in they provide treatment for their clients.

    Convergent evidence validity and discriminant evidence validity are two ways to compare an instrument to another instrument. Convergent evidence validity is when an instrument is related to other variables that should theoretically be positively related. In other words, if I created a new depression inventory I would want there to be high validity coefficient with a well-established depression inventory, such as the Hamilton. The reason I want a high validity coefficient is because that indicates that my new instrument is measuring what it’s supposed to measure, which is depression. Whereas, the discriminant evidenced validity is when an instrument is not correlated with variables that should differentiated from one another. With discriminant evidence validity, we want a low validity coefficient between two instruments measuring two different variables. The reason being is that a low validity coefficient score between two instruments indicates that the two instruments are measuring what they are supposed to. For example, a depression inventory and an anxiety inventory should have a low validity coefficient, because it is measuring two different variables. As Dr. V shared with us, if the new depression inventory that was created had a high validity coefficient with a well-established GAD assessment, that would indicate that there is something wrong with the new depression inventory instrument, and is not valid in accurately assessing for depression.

    Reply

    • Vanessa Nichols
      Sep 21, 2021 @ 12:58:39

      Hi Mary,
      I think you did a great job describing Criterion-related validity. I agree that criterion relation is essential because it helps mental health counselors understand their client’s mental health status. Especially with the case of suicidality because you would never want to send a suicidal client home without discussing it. I also think criterion-related validity is crucial for treatment and discussion purposes. If the assessment does not have criterion-related validity, then any discussion of the results is pointless because it did not measure what it was supposed to.
      Regarding treatment, Suppose a clinician assessed the client’s chance of having a panic attack. It allows the counselor to put in productive factors and develop a game plan before the behaviors even occur. An important part of criterion-related validity is the time between the assessment and the predicted behavior. Good assessment should indicate the expected amount of time between the assessment and predicted behavior.
      I think you did a great job discussing convergent evidence and discriminant evidence. I think the examples you used were clear and to the point.
      Thank you!

      Reply

    • Bekah Riley
      Sep 21, 2021 @ 17:21:05

      Hi Mary,

      I thought your response provided very useful information. First off, I really liked how you defined criterion-related validity in saying that the instrument predicts behavior from the present to the future. In addition, the specific example of a suicidality instrument and how a counselor can interpret the score to infer and predict certain things about the client really helped further my understanding of criterion-related validity in a mental health assessment!

      When describing convergent vs discriminant evidence validity, I also used instruments measuring depression in my examples! I felt as though these examples were very helpful when reading the textbook! I also think you did a great job at defining these different forms of validity as well as going into detail with the examples!

      Reply

  5. Vanessa Nichols
    Sep 21, 2021 @ 12:42:10

    Criterion-related validity, by definition, is the extent that the assessment was related to the outcome criterion. What that really means is how well the assessment predicts outcomes pertaining to the specific criteria. The particular criteria can be anything. In the online lectures, Dr.V talks about suicidality and panic disorder assessment. For my example, I am going to talk about substance use. Let says an assessment exists that can predict substance use behavior. To display criterion-related validity, the assessment must predict clients’ current desire or future use of substances. This is where criterion-related validity gets a little subjective because Criterion-related validity breaks into two subgroups: Concurrent validity and Predictive validity. The only difference between concurrent and predictive is the time lapse between assessment and the predicted behavior or symptom. For concurrent validity, there is no time-lapse. I would give the assessment and then right away talk about potential use. Predictive validity a lapse of time exists (any amount of time). For example, I would give the assessment and ask about substances in two weeks. If the client began using a substance (first time /relapse), then I would say that the assessment displays criterion-related validity and predictive validity. Any good assessment with a time-lapse should tell the time frame in between assessment and the predictive behavior.
    Criterion-related validity is important because it can help the clinician in emergencies (suicide), but more generalized, it helps the clinician with treatment decisions. These assessments can predict potentially harmful behaviors (self or others). If you can trust the assessment, it allows the clinician an opportunity to discuss the potential behavior, put in a coping mechanism, or hospitalization.

    Convergent evidence means the instrument is related (correlation) to other variables that it “should” relate to. For example, if you have an anxiety assessment, you would want this instrument to correlate with other anxiety assessments or specific variables that break down anxiety disorder. Discriminant evidence means that the instrument wouldn’t relate (correlate) with variables it should differ from. For example, if you had a suicidality assessment, you would want to make sure that this assessment measures suicidality and not other variables, even depression. Some people are depressed that are not suicidal, and some people are suicidal, not due to depression. So although they can be related, it’s essential your suicidality assessment only measures suicidality.

    Reply

    • Bekah Riley
      Sep 21, 2021 @ 17:30:30

      Hi Vanessa,

      I thought your post was great! I really liked how you started off by giving a more formal definition of criterion-related validity, but then took it a step further in breaking down that definition in a way that made it easy to understand. I also liked how you went into depth about the subgroups of criterion-related validity. It allowed me to get a better understanding on concurrent and predictive validity.

      I also really liked your description of why criterion-related validity is important in mental health assessments. Being able to trust an assessment provides the clinician with useful information on how to treat clients, and as you mentioned, talk about possible future behaviors that may occur and different coping mechanisms that may be helpful.

      Reply

    • Lauren Pereira
      Sep 22, 2021 @ 09:31:48

      Vanessa,
      I found your definition of criterion-related validity to be very descriptive! It has really helped me grasp this concept even more. You brought great examples into it to make this type of validity more visible. I love your topic of substance abuse that you incorporated in as well. Being able to determine why this type of validity is so important in the mental health field is important and you made a great point in showing that. I agree that it is helpful to the clinician as well as the client, as it is a helpful step within their disorder.
      You also did a great job explaining the difference between convergent and discriminant evidence while still portraying how they can be used together within measuring assessments.

      Reply

    • Tom Mandozzi
      Sep 22, 2021 @ 23:42:14

      Hi Vanessa,

      I really appreciated the points you made in your post as I continue to grasp the concept of criterion-related validity. I agree with your point about how criterion-related validity can support emergency-related treatment. As you mentioned, ensuring this type of validity on a suicidality assessment is very important in treatment for a clinician supporting an individual who is suicidal or is at a high risk for self harm in the future. At the agency I work at, the clinicians I work with are always monitoring and assessing for self harm and safety assessment. Having an assessment that can most accurately predict the risk for unsafe behavior is critical in the implementation of treatment and interventions, as well as supports. By being able to use an assessment with criterion-related validity, we can involve higher levels of care as needed, including psychiatric emergency services, crisis interventions and inpatient services. In the field, safety is extremely important and having this type of validity can support more accurately predicting needs for the individual being served. Thank you for your post – it definitely helped me gain a greater understanding of this!

      Reply

    • Kelsey McGinness
      Sep 26, 2021 @ 09:14:48

      Vanessa I like your discrimination between convergent and discriminant. When you spoke about discriminant needing to only measure the specific target and not other variables that could influence or be related that was clearly put. I had not thought about when measuring suicidality to make sure the assessments was parted away from other variables such as depression as for me they always went hand in hand. Your explanation got me thinking differently, great job!

      Reply

  6. Victoria Cestodio
    Sep 21, 2021 @ 14:33:29

    Criterion- related validity is when we look at the overall instrument and its predictability (especially if it is a good predictor). Dr. V used a great example in the lecture, if someone were to take an instrument how well can it predict future suicidality. This type of validity is very important to our field because as mental health counselors we want to make sure the certain instrument we are giving our client is measuring what we intend it to. For example, if we are trying to measure one of our clients depression, is the instrument predicting potential suicide rates in the future or is it predicting future panic attacks? As you could assume we want it to predict future suicide rates not something that doesnt relate to what we are measuring. When we are working with clients we are not just trying to help them in the present, but we are trying to help their future. I believe this is why this is extremely important to our field. Eventually what we want for our clients is to grow throughout therapy and resolve things (to the best of our ability) that made them come in the first place. We are actively working on our future selves in therapy, therefore measuring predictability is crucial. Dr. V has mentioned the importance of assessments in therapy, such as having a client take an assessment at the beginning of therapy and then taking the same assessment the last session so we measure the progress made.

    Convergent validity takes two instruments that are supposed to measure the same type of construct and shows that they are positively related. An example would be if I wanted to create my own depression inventory. I would then hope that if I compared my own to the BDI it would be positively related because they are measuring the same thing. On the other side there is discriminant validity. Discriminant validity is when two measures are NOT supposed to be related, but rather unrelated. For example, comparing the BDI II to GAD (7). Those two measures are supposed to be unrelated, therefore it would be consistent with discriminant validity.

    Reply

    • Lauren Pereira
      Sep 22, 2021 @ 09:40:49

      Victoria,
      I like how you incorporated Dr. V’s example from his lecture to further explain criterion-related validity. I also like how you explained that this type of validity is important because the certain instrument we are giving our client is measuring what we are intending it to. This type of validity’s main focus is to predict the clients process which is essentially the purpose of the instrument we use.
      I also used the example of Depression when talking about the difference between convergent and discriminant validity. It was interesting to read about BDI being positively related because that wasn’t the first thing that came to my mind but it definitely is an accurate representation for convergent validity within Depression.

      Reply

    • Tom Mandozzi
      Sep 23, 2021 @ 00:04:30

      Hi Victoria,

      I really appreciated that you mentioned the importance of assessing and re-assessing in the context of the mental health field and validity. I work as a therapeutic support in the in-home therapy service line for an agency, and we very heavily emphasize assessment. Every three months, we implement a progress indicator tool, which prompts the youth and their family members to respond to whether they agree with various statements regarding treatment and progress from 1-10 (“1” being strongly disagree and “10” being strongly agree). We are constantly monitoring for progress to inform treatment planning and IAP updates, as well as assessing when services should appropriate close when progress is made.

      Your post made me think about what the criterion-related validity is for this progress indicator tool that we use with our families. How accurately are we predicting progress and future behavior without IHT services moving forward? I think this concept is so important in this field, because it is easy to get stuck in services. Both the client and treating clinician can get stuck in a repetitive cycle where it can be hard to measure progress. By constantly assessing and reassessing, we can truly demonstrate progress and change from start to finish of services and predict future behavior for our clients.

      Reply

    • Tressa Novack
      Sep 23, 2021 @ 14:53:23

      Hi Victoria,
      I like how you pointed out that in therapy our clients are trying to grow and predictability is important because we are concerned about our clients’ futures. I did not say this in my post, but I agree with it. We want to make sure that we can try to predict something dangerous such as suicide. It is a good point that you make when you say that clients are working on their future selves. I think that in therapy we work on our present and past problems in the hopes that our lives will get better in the present moment, but definitely in the future as well. Also, I like the example you give for convergent evidence validity and discriminate evidence validity. Regardless of the fact that depression and anxiety are often co-morbid, it is crucial that instruments measure the disorder they were meant to and that disorder only.
      Tressa

      Reply

  7. Bekah Riley
    Sep 21, 2021 @ 17:09:54

    Criterion-related validity considers whether an instrument is related to an outcome. In other words, this form of validity can be seen as a predictor to a specific outcome. A great example referenced in the textbook that most students who further their education to the collegiate level can relate to is the SAT. The goal of the SAT is to predict how a student will perform in college based on the score that student receives. For example, higher scores in different areas of the SAT will predict that the student will perform well in college, while lower scores predict a lower performance in college. Another similar example is the GRE. This test is designed to predict if undergraduate students can perform at a graduate level. Whether these scores actually predict the performance outcome is the criterion-related validity.

    In mental health assessments, criterion-related validity is especially important. Scores on different mental health assessments may be informative in that treatment would be either very beneficial to a client or not necessarily needed. That is why it is incredibly important for the specific mental health assessment to accurately predict outcomes, or in other words, have a high criterion-related validity.

    Convergent and discriminant evidence validity are very different, yet both very important. Convergent evidence validity refers to how an instrument measuring one thing should relate to another variable or an instrument measuring the same thing. An example of this would be that one instrument measuring depression should have a correlation with another instrument that also measures depression. On the other hand, discriminant evidence validity occurs when variables that should differ are not correlated. An example of this would be that an instrument designed to measure depression will not correlate with an instrument designed to measure anxiety to the point where it is significant. An example of where both of these types of validity would be important is in the creation of a new instrument measuring a specific mental disorder. A new instrument measuring anxiety for example would have to be compared to an instrument that is already well established in measuring anxiety. The convergent evidence validity when comparing the two should be high, while the discriminant evidence validity should be low.

    Reply

    • Emily Barefield
      Sep 22, 2021 @ 20:18:56

      Bekah,

      You did a great job discussing criterion-related validity. I appreciate that the examples that you discussed are something that is familiar to all of us. As you said, the SAT and GRE should be good predictors of future academic performance, this is not always the case, suggesting that these two tests don’t have as much criterion-validity was we would like.

      You brought up a good point that mental health assessments should have high criterion-validity because this informs treatment. We want to give our clients treatments that will be beneficial to them, not ones that will waste their time and resources.

      You did a good job of differentiating between convergent and divergent validity. Another example might be the creation of an instrument to measure OCD. If the instrument is highly correlated with an established measure of OCD, we would say that is evidence of convergent validity. If that same instrument was not highly correlated with an established measure for generalized anxiety disorder, we would say that this is evidence of divergent validity. Great post!

      -Emily

      Reply

    • Pilar
      Sep 25, 2021 @ 23:33:16

      Hey Bekah,
      I like your use of two examples to define criterion related validity, you utilized one from the textbook and then compared it to a second example.

      I totally agree that I’m mental health assessment it is super important to utilize assessments that work to make important discoveries and correct conclusions about conditions. In the field of counseling we want to help not hurt our patients by utilizing incorrect assessments.

      Lastly I find it really helpful how when explaining discriminant versus convergent validity emphasized the comparison of tests that measure similar things so that you can ensure the instrument is effective when describing convergent validity. I also like how you explained how a researcher would want low discriminant validity and high convergent validity.

      Reply

  8. Alexis Grey
    Sep 21, 2021 @ 17:31:52

    Criterion related validity is a component of construct validity which relates to how well a tool or an instrument can confirm or predict a criterion outcome. The criterion is a behavior. Basically, it is how well a tool measures or predicts whatever it is designed to measure or predict in terms of a behavior or an outcome being presented or not presented, to what degree etc. In psychology this means how well a tool is able to predict behaviors like anxiety symptoms, suicidality, depression symptoms and more.

    This type of validity is important in the context of mental health treatment because if a psychologist or counselor is measuring someone for depressive symptoms, or particularly for suicidality and can make a prediction that these symptoms will get worse or not improve, that effects patient care and the level of alarm or fear for client or patient safety we may have. Any time a client is suicidal especially it is imperative that not be ignored. If we are able to predict that someone’s mental state and mental health symptoms are getting worse, then we are better able to form a more effective and aggressive action plan for treatment and intervention. I think another example that stands out to me in terms of being able to better predict is relapse in substance misuse cases. If we are able to predict a relapse, then of course it is beneficial to that client or patient.

    This was a bit confusing to me but… I think…. Convergent and discriminant evidence is based on essentially do correlations exist where they should exist and, do they also not exist where they do not belong. Convergent evidence validity is basically agreement between measures that should measure the same or related concepts while Discriminant evidence validity exists when there is a lack of correlation between the variable in question and other unrelated variables or concepts. The example the book gives is that two instruments which both are both designed to measure depressive symptoms should be positively correlated – said another way the results should be in agreement. While an instrument that is designed to measure levels depression symptoms should not have the same “convergent” relationship or positive correlation with an instrument designed to measure anger, or stress or some other concept. These are different conceptually and so the correlation should be “discriminant”. Two measures should agree if they are similar, and should not if they are not. This is important in mental heath to be sure we are always measuring what we intend to measure and that things like anxiety and depression measures are not becoming so similar that we are combining diagnoses.

    Reply

    • Emily Barefield
      Sep 22, 2021 @ 20:33:59

      Hi Alexis,

      I like your explanation of criterion-related validity and that you discussed how it was a part of construct validity. I also like that your example was related to psychology.

      The examples you gave in your discussion about criterion-related validity and the mental health field highlighted the importance of using instruments that have high criterion really well. Using instruments that have low criterion-validity can have severe, life-threatening consequences, especially in the case of suicide and substance misuse.

      You did a great job of explaining convergent and divergent validity! I know you were a little hesitant but you were absolutely correct. You also bring up a great point about making sure that diagnosis don’t become so similar that they become combined- having high discriminant validity is important for that. Great post!

      Reply

  9. Lauren Pereira
    Sep 22, 2021 @ 08:37:59

    Criterion-Related Validity can also be known as Prediction or Instrument-Criterion. This focuses on how well a measure can predict an outcome for another measure. It will help determine how accurate a specific outcome is, including types of behaviors and emotions that are associated with it. Within mental health assessments, this certain type of validity is so significant because it is being used to predict each clients progress. This determines if the process is going how we expected it to go for our clients and if it is moving in the right direction. Depending on the type of disorder you are dealing with, the criterion related validity can help us predict whether or not there are improvements.

    The major difference between convergent evidence validity and discriminant evidence validity are the items being measured. Within convergent evidence validity, the focus is on measures that are measuring the same constructs, indicating that those two things are related. Discriminant evidence validity involves measures that are not related with one another. It can be confusing to identify the two differences at first because they are both recognized within construct validity. In fact, they actually work well hand in hand. For example, when testing Depression, I would use the Hamilton Scale and hope that it would result in higher validity coefficient since this is a very well-established rating scale for Depression. Next, the key would be to measure something that does not relate so the validity coefficient comes back low. For example, Anxiety Disorder can be the discriminant evidence validity because it will focus on a measure that is unrelated to Depression.

    Reply

    • Madelyn Haas
      Sep 22, 2021 @ 20:45:30

      Hi Laura,

      I enjoyed reading your post. I am glad that you pointed out that convergent and discriminant validity fall under the category of construct validity. I initially thought (and put in my post) that they were a part of criterion validity.

      Other than that, I liked your example for convergent and discriminant validity. It is important to make sure that the instruments we use are actually related to what we want to measure.

      Reply

    • Tressa Novack
      Sep 23, 2021 @ 15:24:10

      Hi Lauren,
      You make a good point that criterion-related validity can be used to predict our patient’s progress. I did not even think of this. Rather, I talked about how it can predict a dangerous event such as suicide. However, pointing out that it can be used to predict progress shows that instruments don’t always have to be used for the negative side of mental health. I also like how you pointed out that convergent evidence validity and discriminate evidence validity work well hand in hand. It did not cross my mind that it would be beneficial to use the two together, but it makes sense when evaluating clients for a certain mental health disorder.
      Tressa

      Reply

    • Teresia
      Sep 25, 2021 @ 18:27:20

      Hi Lauren,
      I really enjoyed your post! I liked your explanation of criterion-related validity and how you brought up that it can help predict improvement. This is not something I thought about. I focused on its use of helping to pick treatment. Great job in pointing out that convergent evidence validity and discriminant evidence validity are both within construct validity. I struggled at first with differentiating the two because they both fell under construct validity.

      Reply

  10. Kelsey McGinness
    Sep 22, 2021 @ 20:28:09

    Criterion-related validity also known as prediction or instrument-criterion is important because its focus is to determine whether an assessment is a good predictor of a certain outcome criterion. For example, this type of validity is good if/when a psychologist is looking to determine the future likelihood of a behavior occurring again in the future. For example, if an individual has a history of stealing and is in therapy as a form of repercussion for repeated offenses. A psychologist may better want to understand the history of the behavior, what motivates the behavior to occur, and any mental or environmental triggers that occur surrounding the behavior and the motivating operant, in order to see if this behavior will continue in the future. Therefore, using criterion-related validity to determine if an assessment best fits the criterion for shop lifting behavior may be the best fit when selecting an assessment.

    Other types of validity include convergent evidence validity and discriminant evidence validity. Convergent evidence validity is when two or more variables are related to an instrument which in theory should then be related/correlated. For example, if a psychologist if looking to assess bipolar disorder on one patient and the assessment is positively correlated to another assessment which measures bipolar disorder then there is in turn convergent evidence. Whereas discriminant evidence validity is stating that two variables that are not related, are in fact unrelated. For example, if you are looking to assess if an impulse control disorder such as kleptophilia is related to other impulse control disorders such as drug addiction and pyromania. Although all are related to the umbrella of impulse control disorders it does not mean that they go hand in hand with one another and are directly related.

    Reply

    • Madelyn Haas
      Sep 22, 2021 @ 20:57:24

      Hi Kelsey,

      I enjoyed your example about the individual with a history of stealing in relation to criterion-related validity. That is not an example I would have come up with, but it makes sense why a practitioner would want to see what predicts the stealing behavior. If there were an instrument with high criterion-related validity on shoplifting, I’m sure it would get a lot of use.

      As for your examples of convergent/discriminant validity, I thought they were both applicable. Convergent and discriminant validity can both be important for the same instrument. For a kleptophilia scale, you would want convergent validity with preexisting kleptophilia scales, but you would want discriminant validity with substance use scales.

      Reply

    • Kristin
      Sep 24, 2021 @ 10:35:42

      Hi Kelsey,

      I agree with Madelyn, I really like your example of someone stealing. My mind immediately goes to more health related risks like suicide obviously, or hurting someone else etc… So I enjoyed reading your post and gaining that perception! Great post!

      -Kristin

      Reply

  11. Madelyn Haas
    Sep 22, 2021 @ 20:34:45

    Validity is a way to see how well an instrument measures what it says it measures. Criterion-related validity is one of the three types of validities along with construct-related and content-related. If I am understanding correctly, criterion-related validity is based on how an instrument relates to an outcome variable. Basically, that means that you compare the scores to the outcome to see how well the measurement works as a predictor. The SATs, for example, should be a predictor of how well students do in college, so they should have a high criterion-related validity. Criterion-related validity is especially important for mental health assessments. Dr. Volungis gave an example of a suicidality scale that I think is especially applicable. If the scale you’ve given your client does not have high criterion-related validity, then you could easily believe that your client probably won’t commit suicide even though they are highly considering it. That makes criterion-related validity in this case a potential life or death situation.

    Within criterion-related validity there are two subtypes: convergent and discriminant validity. Convergent validity means that your instrument correlates with something that you want it to correlate with. For example, if you were making an instrument to measure anxiety, you would want your instrument to correlate with preexisting, well regarded and well researched instruments for measuring anxiety. Discriminant validity is the opposite: It means that your instrument shouldn’t be correlated with something it is unrelated to. For example, if you were making an instrument to measure depression, you wouldn’t want your instrument to be correlated with instruments that measure psychosis. Both types of validity are important, and it is often good to look at an instrument’s convergent and discriminant validities, especially in the mental health field. An example of a measurement where you would want both convergent and discriminant validity would be a depression instrument. You would want your depression instrument to have convergent validity with the Beck Depression Inventory, a well-established instrument. On the other hand, you wouldn’t want your depression inventory to have convergent validity with a Generalized Anxiety Disorder (GAD) questionnaire, but you would want it to have discriminant validity with the GAD questionnaire.

    Reply

    • Sandra Karic
      Sep 25, 2021 @ 16:12:54

      Hi Madelyn,
      I really liked your explanation of criterion-related validity. I think you did a good job emphasizing how poor criterion-related validity could result in a life-or-death situation. I also think you did a great job illustrating the differences between convergent and discriminant evidence. I liked the examples you used and that you related it to a real instrument.

      Reply

  12. Tom Mandozzi
    Sep 22, 2021 @ 23:24:39

    Based on the reading and the lecture, my understanding of criterion-related validity is that it essentially helps to determine an instrument’s ability to predict future behavior. As mentioned in the lecture slides, criterion-related validity is determined to answer the following questions: is the instrument being looked at here a good predictor? Can it appropriately predict one’s behavior in the future? I appreciated the example provided of a score on an instrument that is given to a client and how this score might affect panic disorder symptoms in the future. This is so important in the field of mental health, because ultimately mental health counseling is a form of treatment provided to support improving the symptoms and presentation of mental health disorders. To implement the most appropriate and supportive treatment and to best reduce symptoms of panic disorder, it is critical to develop an instrument that has adequate criterion-related validity in order to predict how symptoms will look in the future and adjust treatment planning and intervention implementation accordingly.

    The example from the book that explained the GRE exam as a predictor of performance and success in graduate school helped me to further understand the concept of criterion-related validity. The criterion involved in the instrument should be reliable in its ability to predict future behavior. I think bias is a huge factor in assessment and a potential roadblock in validity, and I think the book explained this point well with the example of the assessment to measure firefighter effectiveness. There would be a problem if the criterion variable was the firefighter’s supervisor evaluation of job performance, and the supervisors used in the study were biased against female firefighters. In this case, the instrument being implemented would not accurately predict a firefighter’s skills and future job performance. I think similar biases could affect the validity of assessment in mental health as well and further highlights the importance of implementing assessments that have a strong criterion-related validity.

    Convergent evidence validity looks at whether an instrument is related to other variables to which it should be theoretically positively related. At first this was confusing to me, but the example that Dr. Volungis explained in the lecture helped me understand this concept. If Dr. V developed a new depression inventory, it should correlate well with a previously established depression inventory, such as the Beck Depression Inventory. In this case we would want a high validity coefficient to indicate that this new depression inventory is appropriately measuring what it is intended to measure.

    On the other hand, discriminant evidence validity looks at whether an instrument is not correlated with variables from which it should differ. An assessment used to measure depressive symptoms should not be significantly correlated with an assessment used to measure anxiety symptoms and presentation. In this case, we want a low validity coefficient; we don’t want a depression inventory to correlate with a generalized anxiety disorder instrument.

    Another example that I think highlights these concepts would be if I wanted to create a new assessment to predict a student’s performance or readiness for college level classes. The SAT is already a well-established measurement for high school students as a predictor for future college performance. If I created a new instrument to measure this, I would want to make sure that it correlates with the SAT in order to ensure that my new instrument predicts future behavior as intended. This would indicate convergent evidence validity. On the other hand, I would not want this new instrument to correlate with, for example, the GRE assessment. In order to ensure discriminant evidence validity, I would want there to be a low validity coefficient; the test that is meant to measure academic performance in college should not strongly correlate with an assessment meant to measure academic performance at the graduate school level, as these tests are meant to measure different things.

    Reply

  13. Olgena
    Sep 23, 2021 @ 13:25:31

    Referring the textbook and lecture recordings, I think that criterion related validity is a very important element for mental health assessments, because it shows how valid, and well-structured an instrument is, to serve the expected outcome. I believe that the goal of every mental health clinician is to adapt the best assessment/ instrument methods that best fulfill a client needs, and to have future positive outcomes. In addition, while choosing an instrument, it is important to analyze how related and systematic this method is to achieve positive future results. In other words, it is important that this instrument is VALID and a decent predictor on what to expect in the later phases. Listening to Dr. V explanation on this subject, I was thinking about two things:
    One, how important is that the instrument is related with the expected outcome.
    Two, I remembered some of the questionnaires I worked on my undergrad Applied research methods class. Depending on the subject being studied, and the results that were expected from the study, it was important to carefully choose the best survey forms. It was interesting how the questions did not seem direct to the participants (so they could change or think their answers) but were referred to factors that are related to that specific subject. Therefore, based on how well that questionnaire/survey was established, it was a good “guide” to predict the outcome.
    The main difference between convergent evidence validity and discriminant evidence validity is what we are measuring in an instrument, and each of these instruments is measuring different items. As it was described in the lecture, I understood that convergent evidence serves to measure and compare the scores of two different sets of instruments used (two different measures) that are measuring a common set or sets of variables. (Assumed to be related to each other). In contrast, discriminant evidence uses different instruments that are not related, where one’s scores supposed to oppose the scores of the other instrument used. That difference between scores is actually an important element, to show and compare the validity of the instrument used. Both these methods look similar based on the way they are structured; however, they are meant to measure different matters. Based on the examples given in the lecture, if for example I want to measure depression and want to find convergent evidence I must choose a well-established instrument that measures depression. Moreover, to find discriminant evidence I must choose an instrument that is not measuring depression, such as the Stress Scale and Anxiety instrument.

    Reply

    • Monika
      Sep 25, 2021 @ 21:20:47

      Hi Olgena,

      I really enjoyed reading your post, the way you explain the importance of criterion related validity and it’s role in mental health assessment, it’s implications, are necessary to understand as future clinicians. I agree with you that methods for both convergent evidence validity and discriminant evidence validity look similar based on how they are structured but are actually opposite.

      Reply

  14. Tressa Novack
    Sep 23, 2021 @ 14:32:43

    I have to start this post by saying that I am a little bit confused by validity after reading this chapter. I understand that validity tells us whether or not an instrument measures what it is supposed to measure, but all of the different types of validity and topics in the chapter left me a little confused and overwhelmed. From my understanding criterion-related validity tells us whether the instrument used can predict a certain outcome. The book uses the example of the SAT predicting academic performance in college. I think another example of this could be how well the driver’s education written exam predicting an individual’s readiness to drive. This is important in mental health, because if clients’ take instruments that predict levels of depression or suicidality, we want those instruments to be as accurate as possible in their measurement.
    The difference between convergent evidence validity and discriminate evidence validity are that convergent evidence validity shows that an instrument measures what it is supposed to measure, because it has a high validity coefficient with another instrument that is supposed to measure the same thing. Discriminate evidence validity means that an instrument is measuring what it should, because it does not correlate with instruments that measure something else, meaning it has a low validity coefficient. For example, if a clinician wants to determine the severity of their client’s anxiety and gives them two instruments that are designed to measure anxiety, then the two instruments should show similar levels of anxiety severity. If both instruments are showing a similar severity of anxiety in the client, then they have a high validity coefficient and are measuring anxiety accurately. Discriminate evidence is also important here. Let’s say that the client is very anxious, but does not present any other mental health problems. If the clinician gave the client an instrument meant to measure anxiety and an instrument meant to measure suicidality, we would be looking for a low validity coefficient. We would hope to see results that show high anxiety and low suicidality so that we know the instruments are measuring what they are supposed to measure. If we saw high anxiety and high suicidality or low anxiety and low suicidality, we would have to look into other ways to measure our client’s anxiety because they instruments should not be correlating in this way. We would also have to make sure that the client is not suicidal, even though that is not their presenting problem. This example shows why is so important that instruments are making accurate measurements.

    Reply

  15. Monika
    Sep 23, 2021 @ 15:47:35

    Criterion-related validity refers to how well the measurement of one variable can predict the response of another variable. It is an assessment of how well an instrument predicts known related behaviors or constructs. For example, a job applicant takes a performance test during the interview process. If this test accurately predicts how well the employee will perform on the job, the test is said to have criterion validity. There are two ways of assessing criterion-related validity. The first type is concurrent validity where we see if it relates to other outcomes at the same time. The second type predictive validity looks at outcomes in the future. For example, if an individual is scoring high on the extraversion scale then we know from previous research that they are more likely to be outgoing, likely to attend more parties (compared to an introvert), and talk to strangers. Psychological assessment is an important part of both experimental research and clinical treatment. When it comes to mental health assessment, criterion-related validity ensures that the results are an accurate reflection of the dimension undergoing assessment.

    Convergent evidence validity should be able to show a convergence between similar constructs. Discriminant evidence validity should be able to discriminate between dissimilar constructs. For example, if I were to research resilience in college students. To measure resilience(the construct), I use two measurements: The resilience Scale for Adults (RSA) and participant observation. If the scores from two measurements are close enough (i.e. they converge), this demonstrates that they are measuring the same construct. If they don’t converge, this could indicate they are measuring different constructs.

    Reply

    • Alexis Grey
      Sep 25, 2021 @ 19:25:26

      Hi Monika

      I thought your explanation of convergent evidence was super clear, results either converge or discriminate (tell a difference) – I appreciate you articulating that convergence suggests the instruments measure the same construct. For some reason my brain prefers the word “construct” to “criterion”! so this was helpful for me.

      Reply

  16. Jeremy Wronski
    Sep 23, 2021 @ 18:08:41

    Criterion Related validity wants to know whether a particular instrument properly measures the proposed criterion, for example, Weather the SATs measure college success. In a sense, this form of validity is asking whether on not the test can properly predict outcomes. This particular validity is so critical for Psychological tests for a variety of reasons, adherence to theoretical concepts, proper diagnostic power, and having a better understanding of the degree of a client’s symptoms. In the real world, we can use the relationship between the instrument and the criterion to better understand failures in the variability, this is labeled as false positive and false negative knowing these we can identify shortcomings in our test, for example, if a suicide inventory has a high rate of false negatives it would be considered a poor test.

    Convergent Evidence validity is when an instrument is compared to pre-established instruments that measure the same variables to test if the relationship is current.
    Discriminant Evidence validity is when an instrument is compared to pre-established instruments that measure different variables to ensure that the instrument does not correlate with instruments that are not designed to measure the same variables. Notably these two forms of validity measure opposite relationships. Convergent evidence is useful when creating a new test, to ensure that it holds Criterion Related validity. Discriminant evidence could be used when testing a singular individual to see if and result on a test could be better explained by another test, Like if strange depression inventory results could be explained by a present anxiety disorder.

    Reply

    • Moises Chauca
      Sep 24, 2021 @ 19:14:11

      Hello Jeremy,

      I enjoyed your post. I like your examples of false positives and false negatives. They are really important in the mental health field. Clinicians need to know if the assessment have a high rate of false positive or negative. Specifically, since false negatives are terrible because it can lead to the person leaving untreated or pass away. I also like how you explained discriminant and convergent validity and pointed out that they measure opposite variables.

      Reply

    • Monika
      Sep 25, 2021 @ 21:01:16

      Hi Jeremy,

      I really liked how you explained the importance of criterion related validity in mental health assessment. The examples you provided affirms how crucial the results from psychological tests are and their impact on client’s lives. Good point mentioning false positives and false negatives and their impact on test results. Also your point regarding convergent and discriminant evidence validity and how they are opposite is going to help me remember these concepts easily.

      Reply

  17. Sandra Karic
    Sep 23, 2021 @ 18:35:17

    To my understanding criterion related validity is used to evaluate instruments that are meant to predict behavior. There are two types of criterion related validity, concurrent and predictive validity. In concurrent validity there is no time lag between when the instrument is administered and when criterion information is gathered. In predictive validity there is a (significant) time lag between when the instrument is administered and when the criterion information is gathered. I actually could not think of that many examples of potential criterion related instruments used in mental health off the top of my head, but one that I could think of, an instrument that predicts the likelihood someone would commit suicide, would be super important. If the instrument wasn’t valid a clinician runs the risk of not giving this person the help that they desperately need and the person could hurt themselves. 
    With convergent evidence, we examine whether an instrument is positively related to variables that it should theoretically be related to. Discriminant evidence is essentially the opposite, where an instrument with this quality would not correlate with variables that it should theoretically differ from. I think this type of evidence would be important with assessments for mental illnesses, where you’d want the instrument to correlate with instruments that also assess that disorder and not have a correlation with instruments that assess unrelated disorders.

    Reply

    • Alexis Grey
      Sep 25, 2021 @ 19:13:31

      Hi Sandra

      I appreciate you mentioning the time lag – I found this to be initially a little bit confusing and was not seeing a ton of difference in concurrent versus predictive validity – I think when I read the chapter the first time in my head, I had trouble separating the two and also had trouble not connecting concurrent and convergent validity – who names these things – so overall I felt like your response was helpful in clarifying the 4 distinct concepts for me.

      I agree in mental health it is important to be able to be precise when diagnosing someone with any type of disorder – one, because it effects the plan for treatment but also effects how the client may see themselves once a diagnosis or a “label” is placed on them. So many mental disorders share traits and symptoms I can imagine that this line of how similar or how different these things should be in terms of convergent and discriminant evidence is concerned can become very tricky and somewhat subjective for those designing these instruments. The example I am sure comes to everyone’s mind being depression and anxiety measures – similar but not too similar / different but not so different, Certainly interesting to think about the science and judgements behind the instrument design!

      Reply

  18. Pilar B
    Sep 23, 2021 @ 22:10:39

    Criterion-related validity looks at whether or not an instrument is a good predictor of what it’s trying to predict. For example if someone wanted to measure
    work performance for an office job they wouldn’t use an instrument that measures physical fitness it wouldn’t be a good predictor for the criteria however if it’s the goal was to measure work performance for a job like construction then a test of physical fitness would be a more appropriate predictor. Criterion related validity is important for mental health assessments because you want to use a predictor that is appropriate, you wouldn’t use an assessment designed to measure schizophrenia for ptsd. If you have high validity this opens the door to better treatments and interventions for the patients. Low validity creates an uncertainty which you don’t want to have when trying to create a plan of treatment that actually works.

    Convergent evidence validity is when and measure goes well with other tests, because there is evidence that they use the same measurement so if the instrument measures anxiety and another instrument to measure anxiety is correlated to the other one then there is convergent evidence present. Discriminant evidence is when the measure does correlate with another measure of the same thing the variables differ. Discriminant evidence for validity has no correlation with unrelated tests. The two measures account for different variables. So if one measures depression and the other measures OCD then discriminant evidence is present because the variables differ and therefore do not correlate. These two types of validity are important because you want to utilize instruments that are close to other instruments measuring the same thing. Using convergent evidence to create an instrument to assess a mental health condition will ensure higher validity and success of the instrument. If you create an instrument to assess for example ptsd that in no way correlated to other tests use to assess ptsd validity will be low and the results you end up with won’t be very reputable and reliable.

    Reply

    • Alexis Grey
      Sep 25, 2021 @ 18:44:41

      Hi Pilar

      This chapter was sort of needlessly confusing to me and so I really benefitted from your first example of someone wanting to measure work performance – if the criterion being used to predict work performance was physical fitness would not be a good predictor unless it was as you said for a role in construction or maybe working in an agricultural setting or something like that. Putting that real life example of how a criterion can be a good or poor predictor based on context was really helpful for me to understand and to think about this concept. thanks!

      Reply

  19. Jordan Ellis
    Sep 23, 2021 @ 22:26:00

    Criterion-related validity measures how well the instrument is related to an outcome criterion. It is essentially trying to determine how good of a predictor the instrument is. It is important to note that criterion-related validity is looking at the overall instrument, and not individual items. An example of this type of validity would be whether SAT scores predict academic performance in college. Criterion-related validity is important for mental health assessments because it should be able to predict future cognitive and emotional issues. For example, an assessment that measures suicidality needs to be valid because it may be able to prevent future suicide attempts. It also can be predictive of personality types and the development of other mental illnesses.

    Convergent evidence and discriminant evidence basically use the same statistical analyses. The difference is that convergent evidence has high correlation and discriminant evidence has low correlation. During the lecture, Dr. V gave the example of creating his own depression inventory. If his assessment was compared to the Beck’s Depression Inventory, he would want it to be convergent evidence. This would mean that they ask similar questions that accurately measure what they are supposed to measure. Meanwhile, if he were to compare it to the GAD-7, he would want it to be discriminant evidence. This is because they measure two different mental disorders.

    Reply

    • Moises Chauca
      Sep 24, 2021 @ 19:25:37

      Hello Jordan,
      I like your post! Your explanation of criterion-related validity was helpful. I like how you pointed out about the overall instrument measurement and not an individual item measurement. Lastly, your statement about convergent and discriminant validity using same statistical analyses was helpful to understand the similarities. At first, I always thought of them as only different.

      Reply

    • Teresia
      Sep 25, 2021 @ 18:53:16

      Hi Jordan,
      I enjoyed reading your post! You did a great job breaking down the explanation of criterion-related validity. It was helpful for me to remind me that criterion-related validity looks at the overall instrument not the individual item, i think that is an easy point to forget. You also did a great job in explaining convergent evidence and discriminant evidence and clarifying which one has a high and low correlation.

      Reply

  20. Moises Chauca
    Sep 23, 2021 @ 23:32:51

    My understanding of criterion-related validity is how well the overall instrument predicts the outcome criterion. An instrument that has good criterion means it is able to predict what it was designed to measure, or a predicted outcome. For example, in the book, they explain how GRE exams are predictors of how a student will do in college. Criterion-related validity is crucial in mental health assessments because psychological tests needs to be able to predict the intended behavior. A depression test needs to be able to predict if the person has or will have depression symptoms. In addition, criterion-related validity has two important components, concurrent evidence validity and predictive evidence validity. Concurrent validity is when an assessment is able to predict present behavior and there is none or short time length between the assessment and the outcome. Predictive validity is when an assessment is able to predict future behavior and there is a time length between assessment and the behavior. These are really important in the mental health field because clinicians need to know if the assessment is able to predict the behavior and the time length of the examined outcome. An instrument without criterion-related validity is not able to predict an intended outcome and can lead to a patient being misdiagnosed.

    Convergent evidence validity is when an instrument is related to similar instruments that can measure the same outcomes. For example, a new depression assessment should be able to predict the same outcomes as the Hamilton depression scale. The new depression assessment needs to have high convergent evidence with Hamilton so that it is measuring depression. Discriminant evidence validity is a when an instrument is not related to different instruments that measure different outcomes. The new depression assessment should not be able to predict the same outcome as the Hamilton Anxiety Scale because they suppose to measure different outcomes. They are really important in the mental health field because psychological assessment should be able to predict the intended outcome and the same outcomes as their related assessments, so clinicians can identify the issue and provide treatment.

    Reply

    • Jeremy
      Sep 25, 2021 @ 13:23:13

      HI Moises,

      Your breakdown of the books are the example for Criterion Related validity, you broke down criterion-related validity into its parts to better understand the process.
      addidtonlay your explanation of convergent evidence was spot on.

      Reply

    • Sandra Karic
      Sep 25, 2021 @ 16:12:04

      Hi Moises,
      Great job explaining criterion-related validity! I think you did an especially good job in explaining the differences between concurrent and predictive validity and tying it all together by mentioning the possibility of misdiagnosis. I also think you did a good job describing convergent and discriminant evidence. I agree that these types of evidence are important for clinicians who are using assessments to identify issues and treat clients.

      Reply

    • Jordan Ellis
      Sep 26, 2021 @ 06:44:34

      Hi Moises,
      Great job on your post! Thank you for including an explanation of concurrent validity and predictive validity. I can see how important these are in the mental health field. These types of assessments could be used to gain insight on the same issue, but will gather different information. For instance, an assessment with predictive validity would not be used for a client who was at imminent risk of suicide. This is because it would tell us how suicidal the individual might be months or years from now. In this situation, we would need an assessment with concurrent validity, because we need to know how suicidal the individual is in the present time. I appreciate the examples you included for convergent evidence and discriminant evidence. They made it easier for me to understand the difference between the two.

      Reply

  21. Teresia
    Sep 24, 2021 @ 18:06:36

    Criterion-related validity also known as Prediction or Instrument- Criterion refers to how well an instrument validates or predicts the outcome of a particular construct. An example of criterion-related validity would be if a job applicant took a performance test during the onboarding process whether the test would predict the employee’s performance. Criterion-related validity is important during mental health assessment because it helps counselors assess the instrument they are using to make sure that it is actually measuring what they want it too. Giving the counselor confidence in creating a treatment that is best for the client not only right now but also in the future.

    Convergent evidence validity is when an instrument is related to a variable that should hypothetically be positively related. An example would be if a counselor was using an instrument designed to measure depression and the instrument correlates with another instrument that measures depression. Discriminant evidence validity is the opposite of convergent evidence, it is when an instrument is not correlated with variables from unrelated instruments. An example would be an instrument that measures depression is not correlated with an instrument that measures anxiety.

    Reply

    • Jeremy
      Sep 25, 2021 @ 13:27:18

      Hi Teresia,
      The use of prediction in therapy is an incredibly useful tool, we want to make sure our practice is backed by scientific evidence, and if our test has low validity in this field, we can not truly bridge the gap between counseling and theoretical practices.
      Convergent and discriminate evidence is fundamentally two sides of the same coin and can allow therapists to double-check, or critique their tests to ensure proper diagnosis.

      Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Adam M. Volungis, PhD, LMHC

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 72 other followers

%d bloggers like this: