Vignettes: Case Studies and Research at Interview

Read tips here about about how to apply for posts and courses, what to expect in inteviews, in how short-listing is done

Vignettes: Case Studies and Research at Interview

Postby maven » Sat Dec 22, 2007 1:08 am

Case scenario questions, also known as vignettes, are a common component of clinical interviews and test understanding of many issues and ability to apply them in a way that informs practice and research.

Here are some examples to give you an idea of the pertinent topics that can arise. They are presented in the form of example vignettes, followed by a discussion or selection of responses suggested by forum members. These are there to show how different people approach the vignette, rather than as a "right answer" to learn.


Vignette 1


You receive a referral for a young woman, Louise, who is 24. She presented to her G.P. two weeks earlier saying that she feels tired all the time, is having difficulty sleeping, has no motivation and is often tearful. Louise said that she was finding it difficult to be patient with her young son, Thomas (age 2), and often finds herself shouting at him, she worries that she is a bad mother. The G.P. has prescribed antidepressants following a two-week follow up and has referred Louse to you facilitated self help.

- What type of things would you want to know at assessment?
- What do you think is going on?
- What might the difficulties working with Louise be?
- How would you assess risk?

During your assessment session Louise tells you that she is having fleeting suicidal thoughts but that she doesn't want you to tell her G.P. as it is further evidence that she isn't coping and Thomas could get taken into care.

What do you do?

Possible responses 1

I will have to take for granted that I am a trainee and not an assistant, so I'd probably have more freedom to ask her the things below:
First of all, I'd take her short history, past stressful events (looks like she is a single mother), past history of depression or other disorder and prevalence in her family; see if she is receiving any further help from any support groups, parenting groups etc, and formally assess what appears to be moderate to severe depression probably using HADS or something generic like that. I'd also investigate any physical disability and find out more about her sleep patterns. Most importantly I would do a suicide risk assessment by investigating any will for her to take her life and the prevalence of her suicidal thoughts, investigate if she's done any previous suicide attempts, and if she's got a plan and the access to the means to complete it. From what it sounds her likelihood to complete a suicide is not that big as she has a protective factor, her own son. But better be safe than sorry.... If a red flag was raised and she was at high risk, I'd bring it up with supervisor (probably would have done that anyway), investigate the lady's reasons for living and reasons for dying, and link her with short term and long-term resources (friends she can call in hard times, services like Crisis Teams etc). A difficulty that I might have faced with Louise is the complications of her past - it looks like there might have been a sign of abuse in the past, but I would not ask her if she did not disclose such info. So a difficulty would be to avoid getting too deep I guess. I 'd lastly give her some out of hours help line numbers when her suicidal feelings get worse and give her some info on parenting groups etc and depression if it her first episode.

Possible responses 2

That looks pretty nasty to me but a totally reasonable vignette - right here's my attempt at it:
1) what type of thing would you want to know at assessment?
I would ask how long she's felt like this? Is it her first child? How much support she has? (e.g. a partner, friends, family) has there been any significant events that occurred in past year? (e.g. use LES?). You could do the BDI or some other measure such as the HADS.
2) What do you think is going on?
From vignette my 1st assumption would be that she is a young mum (probably 1st child) so feels overwhelmed, and maybe has low levels of support? Her symptoms are suggestive of depression. NICE guidance states medication should only be given after 2-week watchful waiting/follow-up period.
3) What might the difficulties working with Louise be?
I guess 1 problem is that Louise could withhold information for fear of judgment and for fear of losing her child; we don't know what other problems are? I would also consider practicalities such as will she be able to get childcare to attend appointments, if you are doing CBT based interventions will she be able to find time to do homework or be motivated to complete it considering her depression?
4) How would you assess risk?
I'm not sure how to assess her risk, would use some sort of suicide measure? I don't know what though. I would generally ask if she has any suicidal thoughts, plans etc. What prevents her from committing suicide etc? There is the Beck Suicide Inventory but I doubt you would use that in primary care.
5) What do you do?
Regarding the GP issue, I’d think I'd have to inform GP, despite Louise's wish not to. I'd make sure to reassure her that best thing for her child is to be with mother, so she shouldn't worry about losing her child. (Is that correct?). Regarding treatment, I’d provide some sort of CBT based help; get her involved with support groups/social groups to increase her social network; also increase her exercise; and perhaps parenting advice/help. You would of course have told Louise about the limits to confidentiality at the beginning of the assessment session and would remind her about this when she talked about her suicidal thoughts.

Miriam's suggestions regarding above responses

Firstly, I think it would be good to discuss with the client why you would want to tell the GP, and perhaps discuss her fears about this. Hopefully you will have explained the boundaries of confidentiality and the limits to your role when you first met her, and will have some sort of developing trust in your therapeutic relationship.

You could offer some reassurance, but be careful as you don't know the whole situation (maybe social services are already involved for other reasons and there is a real risk of losing the child, or maybe you will later have to raise child protection issues) so be measured in your wording. Perhaps saying that it seems unlikely that she would lose her child if she is otherwise a good mother, especially when she is seeking help. You can also build upon her motivation to change to engage her in therapy ("it sounds like it is really important for you to be a good mum for XXX; Would that mean you'd like to learn to think about things or do things differently?)

In formulating you could see if this appears to be a post-natal depression, or is something she experienced prior to becoming a mother. You might also want to ask if anyone else in her family has depression or mental health issues. You could also ask about the practicalities of her life - who lives in her house, whether she has any regular activities (e.g. a job, regular social opportunities, nursery/playgroup, whether she is in a relationship, or living with family, or alone with her child). Exercise and any activity associated with pleasure or mastery are often good things to recommend in order to enhance mood.

When she reveals thoughts of suicide, it is good to see how often and serious the thoughts are, whether she has a specific plan, what prevents her from acting on the thoughts, whether she is using drugs or alcohol, whether there are particularly risky times for her, and whom she could contact if she felt like that again. You might also want to try and negotiate some sort of contract, where she agrees to use certain strategies before acting on thoughts of self harm, such as using the strategies you have been working on in sessions, ringing the Samaritans (or another help line), contacting her GP or going to A&E (and what she would do with her son in this event). However, it is also good to refer to how you would need to know the local policy, and have discussed these issues with your supervisor, and how you should hopefully be given some more explicit training and guidance, to acknowledge that there is always room to learn more.

Vignette 2 - a research question

You work in a child and family service, which takes referrals of families where a child has behavioural difficulties. The service routinely gathers questionnaire data from parents at the time when a child is referred to the service. One questionnaire measures stress levels in parents. Your supervisor suspects that when parents have higher amounts of stress, it takes longer for the child's behavioural difficulties to improve.
Devise a study that could test your supervisor's belief.

Vignette 3 - research


The government is looking to increase finance in supported employment schemes for people with long term mental health problems. Their scientific advisor believes employment leads to improvements in self-esteem and self-efficacy, and that this in turn leads to reduction in psychiatric symptomology. How would you test this theory?

Vignette 4 - research


You are a psychologist working with medics in an adult asthma clinic. A study conducted by the medics shows that whilst patients use their medication when they get asthma attacks (called "rescue" medication), not all of them use the inhalers that prevent asthma attacks (called "controller" medication). The controller inhalers should be used once every day.

Design a study that investigates whether the difference in use of rescue and controller medication is related to perceptions of the medications’ effectiveness and side effects.

Vignette 5 - a clinical scenario question


What would you do if a client told you that he or she was going to attempt suicide but told you not to tell anybody?

Possible response

I basically said although I respect patient confidentiality, I also have a duty of care to this person therefore would have to tell my line manager about the disclosure. I would explain to the client my responsibility to tell somebody and try to encourage them not to act upon their suicidal ideation.

Vignette 6- a clinical scenario question


What would you do if a patient disclosed that he/she had been sexually abused?

Possible response

I would tell my line manager if I thought that this person (or any other person for that matter) was still at risk of being abused by this person.

Vignette 7 - a clinical scenario question

How would you deal with a client who became aggressive towards you?

Response

Try to de-escalate the situation verbally first - try to talk the person around, use non-confrontational tone-of-voice and body language etc. If the person became physically aggressive, follow workplace procedure, e.g. hit panic buttons, use breakaway techniques, call other staff in to use approved C&R technique.

Miriam's advice
There's nothing wrong with saying something like "I'm not sure, so I would consult my line manager." Some questions ARE looking to see that you would seek appropriate supervision/guidance and also saying this looks much better than waffling an answer that is totally off the mark.

Vignette 8 - research


We have noticed that a disproportionate number of our initial DNA's are from people living on a particular estate. This estate is quite a deprived area, and we hypothesised that this DNA rate is because of difficulty in affording the bus fare. Design a small study to test this hypothesis.

Possible response

We were given 20 minutes to think/write, and then asked about it in the interview. The questions mainly hung round why I had chosen to conduct my study in particular ways rather than other ways, what practical factors I would need to consider, consent, ethics, how many people to include, and then, at the end, how I'd analyse the data.

So for example:

1. Identify 40 referrals over the next 6 months for initial assessment appointments of people whose address is in the estate area. (Using department database/postcodes).

2. Give 20 an appointment letter enclosing the cost of return bus fare, and 20 the regular appointment letter.

3. After all have had appointments compare DNA % rates between the two groups (bus fare/no bus fare) to see if bus fare payment influence decision to attend.

4. Ethical problem. Is it unfair to pay bus fare of half the group and not others?

Some general points regarding vignettes:


You might find it useful to think about what were the general principles behind the questions you were asked - were they about ethics, confidentiality, applying theoretical knowledge? I found that I was more successful at answering these types of questions if I concentrated less on answering specific examples and more on showing that I understood what they were really asking about - it made me more confident about saying things like 'well I've never been in this situation, but in (different but related situation) I did this, and in future I would do that, so applying that knowledge to this situation I would probably do...'

Miriam's advice

I think it's more about showing understanding about what they are asking about e.g. with young children, issues about engagement, issues of consent, etc., and show how you think about these.

The sort of questions could be:
What do you think is going on here?
Who else would you consider contacting?
What sort of intervention would you do?
What are the other issues involved in working with this client group?

As you can see that sort of questions I have come across are very open ended and I suppose to see how you think and express yourself.

Another viewpoint

I agree that questions are often asked to test whether if an issue arose with a client you would try and "handle it" or seek appropriate supervision. For me it wouldn't be appropriate to tell my line manager, it would be my clinical supervisor I take it to and this is something that has cropped up within our Psychology Department and at interviews i.e. appropriate use of supervision. My line manager is responsible for management of my general work - i.e. mandatory training, IPDP, annual leave, fire lectures etc. but she is not a clinical psychologist. If an issue arises with a client I have to take it to clinical supervision as opposed to management supervision. If your line manager is not a clinical psychologist, it's important that you have appropriate clinical supervision and know who to go to, and equally important that you know the difference and are able to demonstrate that you know the difference between different types of supervision. If your line manager and your clinical supervisor are one and the same person, you need to be aware of the potential problems that can arise there and be able to reflect on that too, i.e. where would you go if there was a conflict between the two roles?

Another viewpoint

It's very easy to get wrapped up and panicky about 'official procedures' when presented with a 'scenario, or ethical dilemma. It's good to take a step back and think 'what would I do if I were faced with a situation like that now'. For example, if you are asked a question about 'what to do if a client mentions suicide, or you are concerned about a client', before worrying about official procedures (e.g. duty of care etc), the first steps would be to:

1) Clarification. Talk to the person, encourage them to go into more detail + seek clarification - e.g. are they just suicidal thoughts, or is there suicidal intent? Does the person have plans? If it is a concern that another person (e.g. child) may be at risk, seek clarification on the things the person has said that concern you in a non-confrontational manner.

2) Communication. For a question on suicide risk, I think it would be useful to outline your concerns to the person involved talk to them about what you might have to do + reassure the person whilst talking about what you might have to do if you feel the person is at risk. Similarly, if is about anything else - communicate your concerns + see what they think. This might be more difficult when another person is involved (e.g. child/adult protection issues), as a slightly more covert approach may be needed if you feel another person/party may be at risk.

3) Supervision + note taking. Seek supervision on the issues you have concerns about. Ensure everything you have gone through is documented in the person's notes. Your supervisor should be able to give advice + guidance on what procedures/ steps are or aren't needed. I've probably missed a host of other points, and may be inaccurate in my answer, but that's the way I’d structure an ethical 'scenario' question.

Real Life response to research vignette

At my clinical interview I was given a really brief research proposal and asked to comment on how it could be improved. It was a straightforward sort of thing, I was asked specifically at the end what I thought of the statistical test used - I had no idea but an existing trainee I knew had told me just to confess if I didn't know a research question, so I just said, I used to know about statistics and did well in stats in my degree, but I've forgotten all about stats for the time being and couldn't possibly comment. I got in and still am none the wiser about stats half way through the course!

Another real life response

It seems to me that you wont get a vignette and then get asked "what would you do with this person"(in terms of diagnosis and choice of intervention), it's going to be more along the lines of "how would you monitor/measure treatment outcome" which has more to do with research methods than clinical skill EG:

Two similar research vignettes were outlined. Something about measuring treatment outcomes using a baseline measure, mid-therapy measure and end of therapy measure.
Can you talk about the (dis)advantages of measuring therapy in this way?
How might this benefit / affect the client?
How might it affect the therapeutic alliance?
What would you do if the outcome appeared to be bad?

In terms of directly clinical questions like:

Talk about a client you have worked with and the theory that underpins it.
What were the benefits for this client?
What were the (dis)advantages of this way of working?
What did you learn from the experience?
Did you come across any difficulties / problems whilst working in this way?

Your answer could be "although I have no direct clinical experience the research project that I worked on in the area of x concerned x model. In general terms the benefits for clients would be", or "As I have no direct clinical experience I cannot answer this question in terms of my own experience, however I am interested in x model ..."

And as for the likes of:

Can you talk about a client that elicited some strong emotions in you (possibly negative), and tell us something about the situation you were in?
How did this affect you?
How did you go about trying to deal with it?
What did you learn from the situation?
Has it changed the way that you work?
What did you take from the situation that you might make use of today?

You can relate it to any experience that you have had, e.g. arguments with partner, disagreement with supervisor, fancying someone you shouldn’t etc. You can give strong answers without having had any clinical experiences; you just need to think psychologically and reflectively!

So, is there a "right way" to handle vignettes?

As far as I'm aware, there isn't a formal framework as such. You could approach it from the 'assess, formulate, intervene, evaluate' perspective (which I think is discussed on the BPS website). First off, what information hasn't been provided in the vignette that you would like to find out? Basically, assess. Then, I'd recommend thinking about what the current difficulties are, what started them, what keeps them going, what stops them getting worse, what resources (personal, practical and supportive) are available to the person, what makes them worse, what the person is doing already that helps, and what the person is doing that keeps them going. In other words, formulate! Then, perhaps a quick tour of what a CP might be able to offer, and what issues (power, diversity, practical, professional, legal) might you need to attend to. In other words, intervention! Then perhaps a quick chat about what changes you might expect to see, how quickly, and how would you/the client know when things were better. In other words, evaluate!

What advice can you offer about responding to research vignettes?

The place I would start would be that the best research uses the least complicated methodology and statistics possible to answer the question. However, if you are stuck at the "what research design" bit then maybe it is useful to have a list of the basic research designs:

* Comparing groups (e.g. are redheads more fiery? Do more people in higher socio-economic groups vote conservative?)
* Comparing change over time (e.g. did the government leaflets about HIV increase knowledge in the general population?)
* Comparing change over time in different groups [includes RCT] (e.g. do people who are given omega fatty acids increase their ability to concentrate more than a control group?)
* Looking at associations (e.g. is age related to reaction times?)
* Looking at causal models (e.g. if we measure calorie intake, minutes of exercise, height, gender and base weight, can we predict who will gain or lose weight reliably?)

You need to then think about what is likely to confound the result (sources of bias, things to control for, factors that may impact on both variables and make them appear to be associated even if they are not) to see how you can fine-tune your design with some of these things in mind. Then you can use the flowchart to think about stats.

One thing I tend to do is to try to draw my experimental design, as that makes you think about which groups are involved, when you measure, what you measure, and when any interventions happen.

More advice on research vignettes

I must emphasise this point: there is a lot more to designing a successful piece of research than simply working out the design, and if you're anything like me then these things will all be much more straightforward. For example:

Sampling - whom are you going to use? Where are you going to get them? How are you going to ensure that they would be an ethical group to use, and how would you ensure you'd get informed consent from them? Where will the research take place? How are you going to control for age, gender etc? How many participants do you need? What are you inclusion and exclusion criteria? Etc

Measures - what measures are you going to use? Are they reliable for this population? Are measures available or do you need to create new ones? Do you need objective or subjective measures?

Define your research question accurately - make sure you mention that a thorough literature search will be conducted. Is your hypothesis one or two tailed?

Think about ethics in detail - if people are going to be in a control group for an intervention study, how can you justify this? What will you do to compensate people who are not allocated therapy? Are your selection methods ethical, or could anyone feel pressurised into participating? Ensure you get ethical approval before you proceed.

When it comes to the design, think about the stages of the experiment e.g. if you're measuring something changing over time, what will you need to measure at the beginning, middle and end? The same things or different things? I think if this is a weak spot for you, be honest about that and explain that you would always consult a book (even state which book or books you'd normally use if you like) or a particular person, for advice and support on this matter. It's important to remember that in 'real life' research, decisions are not made in a vacuum, there is a detailed process of consultation and advice-seeking that means you would never, ever be required to come up with a perfect design, on the spot immediately and go with it. Research is always a process; a process of constant refinement and therefore the interviewers will be looking for an awareness of these issues and realities just as much (if not more) than a perfect research design. It was knowing that and believing that that enabled me not only to get through my research interviews, but sail through them. Acknowledging that 'perhaps' I'd do this, but given the time I've had to think about it, I'm not confident it's correct BUT I do know that there are all these other hugely more important things that need to be taken into consideration before I even begin to think about which statistical test I need to use.

What I'm trying to say is this: think about all the other things about the research that are really important, and the design is more likely to naturally follow. But if it doesn't, don't worry - showing your awareness of these other issues will do you far more favours in an interview situation. It is very unlikely that the interviewers would be looking for a 'perfect' design anyway - there are always a multitude of different approaches to one research question. But what they definitely are looking for is how you would approach the question, the kinds of things you'd think about, and why. And I'm certain you can talk more fluently about those things.[/b]

Note: If you have a suggestion about how to improve or add to this wiki please post it here. If you want to discuss this post please post a new thread in the forum. There is information about the structure, rules and copyright of the wiki here.

Content checked by qualified Clinical Psychologist on DATE
Last modified on DATE
Maven.

Wise men talk because they have something to say, fools because they have to say something - Plato
The fool thinks himself to be wise, but the wise man knows himself to be a fool - Shakespeare
User avatar
maven
Site Admin
 
Posts: 1971
Joined: Sat Mar 24, 2007 9:00 pm

Postby elkie » Sun Mar 14, 2010 2:31 pm

In possible responses 1 to vignette 1 it says that Louise is probably a single mum and also that it sounds as if there is history of abuse. I don't see where the evidence is for either of these statements.
elkie
 
Posts: 49
Joined: Fri Apr 25, 2008 6:26 pm

Re: Vignettes: Case Studies and Research at Interview

Postby sv650biker » Tue Apr 20, 2010 10:34 am

maven wrote:
A difficulty that I might have faced with Louise is the complications of her past - it looks like there might have been a sign of abuse in the past, but I would not ask her if she did not disclose such info. So a difficulty would be to avoid getting too deep I guess


Can I ask what sign of abuse there is? Based on the information provided in the vignette, I cannot see any evidence to support this statement (I would even go as far as saying assumption). Did you have additional info in the vignette that you did not provide in the post? I am just curious as to how you came up with this.

Second, you mention that you would not ask louise about the abuse in her past if she did not disclose it initially. Asking about harm from others is something that comes up during a detailed risk assessment. So therefore you would ask about it in that sense. Did you mean that you wouldnt ask it at all or that you wouldnt probe further if her answer is no in all of the 'harm from others' questions?

Look forward to reading your response :)
User avatar
sv650biker
 
Posts: 66
Joined: Fri Jun 13, 2008 2:30 pm

Response to Vignette 2.

Postby sb123 » Mon Nov 29, 2010 2:16 pm

Hi,

Nobody seemed to write a response for vignette 2 so I wondered whether my answer would be along the right lines...

The study hypothesis would be that there is a significant difference in the length of time taken for the child's behavioural difficulties to improve between parents experiencing high and low levels of stress at the time of referral.

The data for the parents level of stress could be gained from the questionnaire i.e. a rating scale.

The data for length of time could be calculated by looking at the number of days taken for the childs behaviour to improve - this could be assessed through direct observation and noting down when change has happened and is being maintained.

In terms of statistics, an independent t-test could be done where the IV would be the parent group (high/low stress level) and the DV would be number of days taken for the child's behavioural difficulty to improve.
A pearson's correlation could also be performed to see if there is a relationship between stress level and length of time - however this doesn't infer causality.

Also, parents could be interviewed to see whether they feel their stress levels have an impact upon their children's progress. It would also be interesting to see what kind of stressors are experienced.

Any feedback for this answer is greatly appreciated

Sb :)
sb123
 
Posts: 117
Joined: Thu Sep 13, 2007 5:22 pm

Postby miriam » Mon Nov 29, 2010 9:25 pm

Seems fine to me.

As to the above comments, I can't remember exactly where this was compiled from, livechat or forum discussions, but I think people were wondering why she was so worried about shouting at her child and thinking he would be removed into care - so some people had gone down the path of wondering if she had done more than shout, or if she was trying to avoid experiences she had had herself in childhood, perhaps of being blamed, hurt or removed into Care. I think the assumption of single mum came from no mention of a partner in the referral. These wondering can help ensure that you've gathered enough information in the assessment to rule out such hypotheses.
Miriam

See my blog at http://clinpsyeye.wordpress.com
User avatar
miriam
Site Admin
 
Posts: 7330
Joined: Sat Mar 24, 2007 11:20 pm
Location: Bucks

Re: Response to Vignette 2.

Postby katz » Mon Mar 28, 2011 4:27 pm

sb123 wrote:Hi,

Nobody seemed to write a response for vignette 2 so I wondered whether my answer would be along the right lines...

The study hypothesis would be that there is a significant difference in the length of time taken for the child's behavioural difficulties to improve between parents experiencing high and low levels of stress at the time of referral.

The data for the parents level of stress could be gained from the questionnaire i.e. a rating scale.

The data for length of time could be calculated by looking at the number of days taken for the childs behaviour to improve - this could be assessed through direct observation and noting down when change has happened and is being maintained.

In terms of statistics, an independent t-test could be done where the IV would be the parent group (high/low stress level) and the DV would be number of days taken for the child's behavioural difficulty to improve.
A pearson's correlation could also be performed to see if there is a relationship between stress level and length of time - however this doesn't infer causality.

Also, parents could be interviewed to see whether they feel their stress levels have an impact upon their children's progress. It would also be interesting to see what kind of stressors are experienced.

Any feedback for this answer is greatly appreciated

Sb :)


Hi Sb,

I have just been re-visiting this thread in light of impending interviews and thought that i would add to your comments on this vignette.

Without detailing my full answer, here was a summary of my thoughts...
I too felt that the use of existing data was a good point to start at.

I decided to look at a snapshot of data from the last 50-60 families discharged (due to improvements in behaviour)
I felt that initial investigations to confirm the correlation between parental stress and improvement was necessary. DV: 'stress score' derived from the questionnaire to measure stress, and 'number of sessions' while in therapy as an indicator of the length of time/ degree of input necessary to see sufficient change to discharge.
Confounding variables: I also felt that it was important to consider the possible extraneous variables and therefore would want to also capture some basic demographic data including ethnicity and age (others would assume what data was routinely collected).

I began with a Pearson correlation to assess the relationship between the extraneous variables and 'number of sessions'. assuming no relationship was identified I then progressed to a Pearson correlation comparing our two dependent variables.

Form this point however I picked up on your comment regarding causation. You said that correlational analysis would not infer causality (from which i took to mean this was your reason for conducting a t-test). My understanding is that it is not the statistical method but the experimental methodology that determines our ability to infer causation.

Therefore my next step (assuming a positive correlation had been identified) was to devise an experimental method. My initial thoughts were that we could identify all those who enter the service scoring above a specific threshold for stress indicating that there was a high level of stress. Half of these families would first be offered an intervention to reduce stress levels (and this assessed on the same measure) and half would not (control group). From this point on the intervention would be the same for both groups.

I would then conduct an independent samples t-test to identify any significant differences between the number of sessions attended for the high and low stress groups. As both groups were originally identified as 'stressed' and one group had stress levels reduced prior to intervention, any difference observed can be assumed to be as a result of 'stress' (and the intervention) rather than another confounding variable.

There are clear practical difficulties with this including the identification of a method of stress reduction that is quick (to allow for therapy to commence quickly) and effective (to reduce stress levels). Where the childs behaviour is the cause of the stress this may prove challenging. There are also ethical issues of using data collected from discharged patients....

I don't believe this is the best design due to these limitations but i was struggling to arrive at an alternative experimental design.

any comments on this would be welcome.

Katz
User avatar
katz
 
Posts: 459
Joined: Fri Jul 03, 2009 12:52 pm
Location: Midlands

Re: Vignettes: Case Studies and Research at Interview

Postby Qwerty » Sat Apr 21, 2012 3:17 pm

I think that there are also ethical considerations about not offering some clients stress management techniques if it has already been found (through the Pearsons correlation) that there is a positive relationship between parental stress and disruptive beahviour. For this reason, do you think it would be more ethical to leave it at the Pearsons correlation and not design the experimental investigation? Obviously you would have issues re causality but I think they are less significant than the issues regarding ethics. What do you think?
Qwerty
 
Posts: 91
Joined: Fri Nov 18, 2011 1:55 pm

Re: Vignettes: Case Studies and Research at Interview

Postby miriam » Sat Apr 21, 2012 4:49 pm

A positive correlation means higher stress = higher disruptive behaviour, therefore why would reducing stress not be ethical?
Miriam

See my blog at http://clinpsyeye.wordpress.com
User avatar
miriam
Site Admin
 
Posts: 7330
Joined: Sat Mar 24, 2007 11:20 pm
Location: Bucks

Re: Vignettes: Case Studies and Research at Interview

Postby Qwerty » Sat Apr 21, 2012 6:24 pm

Because you are not giving 50% of the participants access to stress reduction help and denying help is unethical
Qwerty
 
Posts: 91
Joined: Fri Nov 18, 2011 1:55 pm

Re: Vignettes: Case Studies and Research at Interview

Postby sarahg » Sat Apr 21, 2012 9:27 pm

That is often done within research actually, especially when testing drugs.
As long as after the experiment the control group are offered a choice of receiving the stress reduction help, this then meets ethical considerations.
User avatar
sarahg
 
Posts: 614
Joined: Sat Dec 10, 2011 1:46 am

Re: Vignettes: Case Studies and Research at Interview

Postby miriam » Sat Apr 21, 2012 10:45 pm

A correlation is not the same as causation. In this example higher parental stress is associated with more disruptive behaviour, but we don't know which caused the other or if they were both caused by some other factor (eg both being a response to parental unemployment or domestic violence or substance use). If you do an RCT in which you manipulate one of the factors, this doesn't have a predetermined outcome on the other variable.

The other reason is that doing nothing is the best point of comparison. Not every family with parental stress and disruptive child behaviour would seek an intervention, and of those that seek an intervention nationally, not all will be referred to the service offering this intervention, and of those referred some will have to wait.

In that context, I'd argue it is absolutely ethical to have a waiting list control or a no treatment control or an alternative treatment control.
Miriam

See my blog at http://clinpsyeye.wordpress.com
User avatar
miriam
Site Admin
 
Posts: 7330
Joined: Sat Mar 24, 2007 11:20 pm
Location: Bucks

Re: Vignettes: Case Studies and Research at Interview

Postby Qwerty » Sun Apr 22, 2012 12:33 am

I understand your rationale but it would still make me feel uncomfortable about having a no treatment control in this case. I guess people think differently about this sort of thing.
Qwerty
 
Posts: 91
Joined: Fri Nov 18, 2011 1:55 pm

Re: Vignettes: Case Studies and Research at Interview

Postby miriam » Sun Apr 22, 2012 1:12 am

But how would the knowledge base ever progress if you just left it at the correlation stage? Even with a promising new cancer drug there has to be an RCT to compare it to placebo and/or existing treatments before it can be established if it really works and should be used more widely...
Miriam

See my blog at http://clinpsyeye.wordpress.com
User avatar
miriam
Site Admin
 
Posts: 7330
Joined: Sat Mar 24, 2007 11:20 pm
Location: Bucks

Re: Vignettes: Case Studies and Research at Interview

Postby Borrowed Cone » Sun Apr 22, 2012 1:03 pm

AliMartha wrote:Because you are not giving 50% of the participants access to stress reduction help and denying help is unethical


I think you are looking at it the wrong way round. The fact is, half of the participants are not receiving something that they wouldn't be receiving anyway. So it's not like you are depriving them of a service. You are exploring whether a particular "treatment" works or not and comparing that to a group of people who would be doing exactly the same thing as if the research wasn't happening.

As Miriam has pointed out, correlation does not equal causation, so a purely correlational design would ultimately be more of a waste of participants' time than doing a proper controlled study - that is more of an ethical issue, I think.
Last edited by Campion on Sun Apr 22, 2012 4:48 pm, edited 1 time in total.
Reason: To fix the BBCode.
User avatar
Borrowed Cone
 
Posts: 1269
Joined: Fri Nov 12, 2010 11:05 pm
Location: M25


Return to Wiki: Applications, short-listing, interviews, and gaining posts

Who is online

Users browsing this forum: No registered users and 1 guest