Clinical Methods Flashcards

Question

strengths of meta analysis

Answer 1

An advantage of meta-analysis is that it is a way of finding out trends about an issue and any relationships that might exist. A meta-analysis involves statistical analysis and because it uses more data than single studies (the meta-analysis combines the results of studies) then the power of the statistical result is larger than the result of the single studies. Smaller samples in studies become a large sample in a meta-analysis, so there can be better generalisability of findings. Also, studies may be in different cultures, which might enable the meta-analysis to draw conclusions and to generalise about the universality of an issue (if the results from the different cultures relate to one another). A meta-analysis can help to uncover patterns that can then be the subject of further research. This can help in building a body of knowledge.

Answer 2

A difficulty is that the studies that the meta-analysis draws on are unlikely to be identical in their research method, procedure, sampling and decision-making. Putting the results of different studies together needs careful decision-making to make sure that the data that are compared are fairly comparable. Stafford et al. (2015) had to discard results in order to choose just the data that could be compared, for example. Badly designed studies that are used in a meta-analysis lead to a poor meta-analysis, which is why Stafford et al. (2015) chose to grade the quality of the studies they used. It can be argued that if there is doubt about the quality of a study, such as there being bias, then that study, if used, is going to give bias to the results of the meta-analysis. A meta-analysis will draw on studies that have been published in an area of study, and this means unpublished studies of course will not be used. This might show publication bias and distort the findings of the meta-analysis. Studies that show negative or non-significant results are less likely to be published. Ferguson and Brannick (2011) found in a review of meta-analysis studies, that approximately 25% of showed a publication bias issue.

Answer 3

Recap of Carlsson et al. (2003) Carlsson et al. (2003) conducted a review (meta-analysis) of existing research on schizophrenia, focusing on the role of neurotransmitters like dopamine and glutamate. The study analyzed findings from various sources, including brain scans and drug trials, to challenge the traditional dopamine hypothesis and suggest a glutamate-dopamine interaction as key in schizophrenia. Evaluation of Meta-Analysis in Clinical Psychology (Using Carlsson et al., 2003) Strengths: Increased Reliability – By reviewing multiple studies, Carlsson could identify consistent patterns, strengthening the validity of neurotransmitter theories in schizophrenia. Time and Cost Efficient – Instead of conducting new experiments, Carlsson used existing data, avoiding the time and ethical issues of working with vulnerable patients. Reduces Bias of Single Studies – Using diverse data sources minimizes the risk of individual study errors affecting the overall conclusions. Ethical Advantage – Avoids exposing new participants to potentially harmful drug tests or invasive brain scans. Weaknesses: Data Quality Issues – The reliability of a meta-analysis depends on the quality of the original studies; if the studies used have flaws, the conclusions may be unreliable. Lack of Control Over Variables – Carlsson relied on different studies with varying methodologies, making it harder to ensure consistency in findings. Potential Researcher Bias – Selecting which studies to include could introduce bias, as researchers may favor studies that support their hypothesis. Findings May Become Outdated – As new research emerges, previous meta-analyses may become less relevant, requiring continuous updates. Conclusion Meta-analysis, as used by Carlsson et al. (2003), is a powerful tool in clinical psychology, allowing for a broader, more ethical, and reliable understanding of schizophrenia. However, its effectiveness depends on the quality and consistency of the original studies, making careful selection and interpretation essential.

Answer 4

Case studies are interested in individuals and in detail about them. They can be said to have an idiographic approach, which means focusing in detail on a topic and on individuals. Case studies are used to find out detail about something rather than for building cause and effect understanding. They describe someone or a small group as much as they explain them. Case studies are called a ‘research method’ here, though they are not quite the same as other research methods, like experiments. This is because case studies use other research methods like interview observation or questionnaire to gather data about the case they are interested in. They gather mainly qualitative data, because such data give the richness and detail that a case study is looking for. A case study will often use more than one way of collecting data, such as questionnaires for people who know the individual being focused on, or experiments on brain damaged patients, for instance, as well as observations and interviewing to find out more. This means that a lot of the data are qualitative but there can be quantitative data too. Triangulation can be used to look for reliability and validity. Triangulation means taking data from at least two different sources and checking whether the data agree. If the same data is gathered using different methods, then the data has retest reliability. The data are also likely to be valid, because if data match and some data at least come directly from an individual, then the data are likely to be measuring what they claim to measure, which gives the data validity. The process of triangulation add credibility to the results of a study. Nomothetic - This approach focuses on large groups of people and seeks to establish general laws that can be applied across a population. Nomothetic methods include experiments, surveys, and statistical analysis. Idiographic - This approach focuses on the individual and seeks to understand their unique characteristics. Idiographic methods include case studies, interviews, and observations.

Answer 5

Case studies are good at finding out detailed and in-depth information about an individual or small group, and such detail means rich data to draw conclusions from. The advantage here is that the case study allows depth, which is hard to find using other research methods. Experiments and surveys, for example, gather focused data, looking at a specific variable rather than wider and richer data focused on the whole individual or group. Case studies allow that richness, variety, and detail, where other methods often do not. Case studies can look at rare or unique conditions, where samples of such individuals are by definition going to be difficult to find. If a wide sample of a certain type of person is not easily available, meaning data cannot be collected in a more general way, a case study is a way of gathering valuable insights about a rare occurrence. For example, if someone has specific brain damage, then knowing a lot about that case can be crucial in understanding the effects of brain injuries and improving treatment options. As case studies tend to use more than one way of collecting data, triangulation can be used to test for both reliability and validity and also to show consistency in the findings.

Answer 6

A main difficulty is that the focus on a single individual or small group means that generalizing the findings to others is not really possible. The sample is so limited it is not going to reflect the diversity of the wider population, and so results may be limited in their usefulness. Another limitation is that a case study is often carried out by one researcher (such as Freud’s (1909) Little Hans) or a small team, who will get to know the individual who is the focus of the case study very well. This might not cause overt bias, but it is possible that there is subjectivity in the data gathered, and the findings might be affected by the researcher’s input. This is especially the case in clinical psychology, whereby many diagnoses and conclusions are influenced by the individual interpretations of the researcher, leading to potential bias in understanding complex psychological conditions.

Answer 7

An interview is a conversation between two or more people where questions are asked to obtain information about the interviewee. Interviews can be divided into two rough types, interviews of assessment and interviews for information. Interviews involve face-to-face situations or telephone contacts in which the researcher orally solicits responses. The interviewer will find out about the personal data required for the study, such as gender, age, marital status, employment status and other details of relevance. The interviewer will use some standardised instructions at the start of the interview so that the respondent is aware of ethical issues such as confidentiality and the right to withdraw. The respondent will also be told something about the interview and what the purpose is. Interview schedules must be planned out well in advance so that all areas are fully operationalised and an aim or hypothesis formulated. The way of recording the data can be either recorded or written down, whatever format the full conversation it must be fully TRANSCRIBED after the interview. This can be very time consuming in many ways. Goldstein (1988) looked at differences in how males and females experience schizophrenia, using both primary and secondary data from interviews. Goldstein used trained interviewers to gather data about the symptoms of the patients and she used questionnaires administered by an interviewer to gather information about their past histories, family structure and previous experiences.

Answer 8

Structured Interviews Definition: These are highly controlled interviews where the researcher asks a set list of predefined questions in a specific order. Advantages: Easy to replicate, reliable, and quick to analyze as responses are standardized. Disadvantages: Lack of flexibility, may limit the depth of answers, and the interviewee has less freedom to express themselves. Unstructured Interviews Definition: These are more flexible and conversational, with no set list of questions. The interviewer may have a general topic in mind, but the conversation can flow freely. Advantages: More in-depth and rich responses, allowing the researcher to explore new topics as they arise. Disadvantages: Time-consuming, harder to analyze, and results can be influenced by interviewer bias. Semi-Structured Interviews Definition: These combine elements of both structured and unstructured interviews. The interviewer has a list of key questions but can follow up on responses and ask additional questions. Advantages: Balances structure with flexibility, allowing for depth while still covering specific topics. Disadvantages: Can be inconsistent across interviews, making analysis more difficult than structured interviews.

Answer 9

Unstructured interviews are useful for obtaining qualitative data because there can be exploration of the issues and the respondent is able to use their own words and ideas. The qualitative data offered by interviews can give more detail, depth and validity that quantitative data. This may be beneficial in clinical psychology when looking at issues such as the experiences of mental health stigma, effectiveness of treatments, or how an individual functions in their wider social circumstances. Data tends to be more valid in interviews than questionnaires because the respondent can use their own words and issues can be explored face-to-face (or telephone) with the interviewer. This represents the experiences of mental health issues much more realistically and allows the individuals to expand on their points. Interviews enable a large amount of data to be collected which is descriptive and may give a better picture of what is going on in real life so are valid to what is being studied. Interviews give access to information which is not available through direct observation, such as what individuals think and feel about certain topics which again makes it a more valid method.

Answer 10

The interviewer may affect the findings because of the way questions are asked, the way he or she is dressed or other characteristics. This is known as interviewer bias. If an individual with a mental health problem is being interviewed they may feel uncomfortable or nervous and not give reliable responses. There might be subjectivity involved in analysing interviews. When categories and themes have to be identified from in-depth and detailed data, usually involving a transcript, the researcher may allow personal judgements and experiences to affect their analysis. This links to the concept of clinical interviews for diagnosis, where the judgements of subjective concepts, like failure to function adequately, are determined by clinicians with their own social norms, values and beliefs. Objectivity is required to build a body of knowledge, so subjectivity must be avoided where possible in research.

Answer 11

Thematic analysis involves looking at qualitative data and picking out themes and patterns in the data.. Themes are the patterns and they help to describe the overall data, often in relation to a specific research question. Themes/patterns are identified from coding, they then become categories and the data can be examined carefully, allocating the data to the different categories. Thematic analysis takes place by: ● familiarising oneself with the data ● generating codes ● looking for themes in the codes ● reviewing the themes ● defining and naming the themes ● producing the report. The researcher must be very familiar with the data in order to generate the codes and also then to identify themes in the codes that suit the particular research question. Thematic analysis involves generating codes, which can come from theory, from how the researcher understands their data or from previous research. Thematic analysis is a main tool for analysing qualitative data and can be used in other ways to analyse qualitative data, including grounded theory.

Answer 12

Issues of thematic analysis Issues of thematic analysis include how ‘long’ a pattern or theme should be, so that it is put into a category and coded. A pattern would tend to suggest there are quite a few instances of that point or theme, and that it would be quite a large theme, so that a lot of the data would fit into it. That would be efficient in terms of reducing the size of the qualitative data. However, it is not about counting the instances a pattern arises or how much of the data would fit into that pattern. It is not about bringing quantity into the data. A pattern that did not appear often but was judged to be a theme would be accepted by a researcher. The researcher would judge whether an idea is a theme according to the research question and importance of that pattern. A researcher would look for data that captured something in relation to the research question, rather than for ideas that appeared ‘most often’, for example. This would be thematic analysis driven by theory. A researcher might want to capture all the data that has been collected, in which case themes are more likely to be ideas or features of the data that appear often. This would be thematic analysis driven by the data (and more like grounded theory). There are no rules about how large a theme must be to be a ‘theme’. Content analysis involves categorising qualitative data and counting instances of categories. Content analysis is very similar to thematic analysis but is more likely to involve quantity (such as frequency of an idea or feature in the data). Lim et al. (2014) used thematic analysis of the evidence found in other studies looking at placement (work) and support as an intervention for those with schizophrenia or schizoaffective disorder. They say they used a literature search using the internet and found 358 articles. Lim et al. (2014) followed steps suggested by Braun and Clarke (2006). The steps were familiarisation with the data, generating initial codes and then building and reviewing themes. To do the coding they coded each study ‘line by line, from the introduction to the conclusion’. They used 627 excerpts from the data and had 417 codes initially. Then using thematic analysis of the codes and grouping excerpts, checking the data and going back to the themes, they put the 417 codes into 40 categories. The first author did the initial coding, then the second author did the initial coding (separately) and the third author reviewed so that there was consensus. Themes and patterns came from the codes. The researchers follow the steps put forward for doing thematic analysis, such as starting with familiarisation with the data and then coding carefully before building themes. The separate authors act as coders so that reliability can be shown. Themes and patterns come from the coding, which is how thematic analysis works.

Answer 13

● Thematic analysis is a flexible way of analysing qualitative data. It can be done using categories and themes that come from theory, or it can be used to generate categories and themes from data directly. Thematic analysis does not need a theory to drive the analysis and, unlike grounded theory, it does not need to derive a theory from the data; it can do either. ● Another strength is that it is a way of maintaining a richness in data and yet summarising a large amount of qualitative data in a manageable way. ● Thematic analysis allows insights that are not anticipated by the researcher and presents correct and detailed data in a way that does not lose its richness. Participants can be collaborators, for example, and can verify the data once they are analysed, which maintains richness. Participants do not have to be trained to understand the data after the analysis; the data are accessible to everyone as they are written in a clear format, again maintaining the richness and detail.

Answer 14

● A weakness is that it is hard to say how the analysis is done in general terms. There is a tendency to think of themes emerging from the data, which in a way they do, but if a researcher sees that their data show repetition about a certain phrase or feature when many interviews are considered, then that phrase or feature seems to ‘emerge’, but it’s because the researcher has noticed the feature or phrase. Ely et al. (1997, cited in Braun and Clarke, 2006) say it is thinking about the data and creating links that gives the themes; they do not ‘reside in the data’. ● If data come from a questionnaire or an interview, they come from questions put to the participants, and that is often the case. It can be hard to find themes in the data that are not driven by the questions in the questionnaire or interview, as that has already guided the data offered. ● Thematic analysis can be said to have subjectivity. Those analysing the data are likely to make judgements, for example, about what makes data form into a ‘theme’. Science is about objectivity so that there is no bias in data. However, using more than one researcher to analyse the data can reduce subjectivity.

Answer 15

Grounded theory is a way of analysing qualitative data and means not using existing theory but finding theory from the data – the theory is grounded in the data. The goal of analysis is to generate a useful theory from the data – a theory that explains that data. Grounded theory is a research method as much as a method of analysing data, because it focuses away from the positivist tradition of ‘doing science’. Science is about taking a theory, deriving a hypothesis from the theory, testing the data empirically (from the senses) and then accepting, adapting or rejecting the theory because of what was found. Grounded theory focuses in the opposite direction and avoids the adoption of a theory before looking at the data; it involves looking at the data first before arriving at a theory. The researcher using grounded theory looks at the data, sees themes and ideas that are repeated, uses them as categories and groups the data accordingly. After that a theory is put forward. Note that when grouping the data and using coding, grounded theory is no different from thematic analysis. It is just that grounded theory is a method that does not analyse data with a theory already in mind.

Answer 16

How grounded theory takes place: - First, coding is done. Coding involves taking text in small pieces and putting the text into a heading that suits it. - Second, codes are collected into concepts, which groups the data together. The idea is to see how the coded data might fit together and in a way to generate concepts that are ‘wider’. - The theory comes from collecting the categories together; it is about forming a model that can explain the data. Grounded theory uses both inductive and deductive thinking: ● Deductive thinking refers to going from a theory to ‘deduce’ what will arise from that theory. An example is ‘if all men are mortal, and Henry is a man, then, by deduction, Henry is mortal’. It would be the job of the researcher to check the claims, and if the claims are true then the conclusion must be true. ● Inductive thinking refers to the ideas arising from empirical evidence. The idea is empirical observations are carried out and then a general theory is derived. Grounded theory focuses on inductive reasoning but there is deductive reasoning too. In grounded theory the data drive the concepts and from the concepts come the theory, which uses induction. However, checking back from a possible theory to the concepts and back again to build the theory can involve deduction. Grounded theory is about developing a theory from the data – it involves explaining the data, not simply describing it. Coldwell et al. (2011) looked at how people with the diagnosis of psychosis or schizophrenia might contribute to their family. They interviewed six people diagnosed with schizophrenia and six who had a relative with schizophrenia, and then analysed their data using grounded theory. The researchers kept a reflexive journal and memos and analytic auditing was done, so that the reliability of the analysis could be checked. As themes began to emerge they were explored using follow-up interviewing with the participants so that there was more understanding of the categories that were emerging from the data. This is the process of deriving theory that is grounded in data.

Answer 17

● There is validity in that the coding is done carefully using meaning from the data in chunks, and then codes are developed into wider concepts, going backwards and forwards from the codes to the concepts in order to build the concepts. This means that the participants’ own thoughts and feelings are used to drive the analysis, which means the concepts should measure what they claim to measure, which gives validity. ● Grounded theory uses specific terms to explain how it is done and is explicit in its guidance, which, when it comes to deriving theory from qualitative data, is not done in any other way. Thematic analysis can analyse qualitative data and content analysis also can analyse qualitative data and give categories, but it is grounded theory that then goes on to develop a theory to explain the data. Grounded theory allows the creativity that is needed to move from data about people to draw up ideas about how to explain that data. ● Another strength is that using grounded theory allows the richness and detail of the qualitative data to survive the analysis. The concepts that come from the coding bring the richness of the data to the analysis, rather than losing the detail.

Answer 18

● It is perhaps impossible to code and categorise data into concepts without some theory in mind. We cannot put out of mind theories when analysing qualitative data. ● Another similar criticism is that the search is for the shared meanings of individuals that have given the data and yet the way grounded theory is carried out has a formula and is prescribed. This means it would appear to go against the idea of encapsulating rich meaningful data in the meanings of the participants (rather than those of the researcher(s)). ● It could be argued that it is not appropriate to ignore previous research to aim to generate a completely new model or theory from empirical data. The original question is likely to have come from previous research and previous findings, probably including previous theory. ● Engaged theory is another idea. It focuses on generating concepts from data, as does grounded theory, but engaged theory uses existing theory to help in the analysis. ● It is not possible perhaps to generalise from the findings of a study using grounded theory to say that any theory that comes from that data can explain all examples of what is being understood. The data will come from a specific culture in a specific time-frame, using specific participants. Generalising is, therefore, limited.

Answer 19

Measures of central tendency: Where is the middle of your data? Mode. This tells you what the most common score is. To work it out, count how many times each value appears in your data. The most common is your mode. If two values appear the same number of times, then you have two modes. The mode is most useful when you have data that is categorical or ordinal, for example if your data is ‘yes/no’ questions. Median. To work this out, put your data in order, and count half way through. The value right in the middle is your median. If two values are in the middle, work out what is half way between them: that is your median. This is a particularly useful statistic if you have extreme scores. Mean. To work this out, add up the data and divide it by the number of values you have. For example, if ten participants gave you a speed estimate, add up the ten estimates and divide this by ten. This is the most common way to describe the middle of your data set. Measures of dispersion: How spread out are your data? Range. You can easily work out the range of your data by subtracting the bottom score from the top score. This is a very basic way to describe the spread of your data. Standard deviation. This is usually more informative. The standard deviation is calculated based on the average distance from the mean of your data: if you have a lot of extreme scores, you will get a higher standard deviation; if your data are very stable and clustered closely around the mean, you will get a lower standard deviation. What is most useful for you to look at is the difference between the standard deviations of the groups in your study. If there is a big difference in the standard deviation, this means one group has more variation in their scores.

Answer 20

Chi-squared To test the significance of the association/relationship. · To test if two categorical variables are linked to one another · It can be used with nominal data · It can be used with data that does not have normal distribution · The data must only fit ONE category of the contingency table Spearman’s rank correlation co-efficient To test the significance of a correlation. · To see if there is a correlation between two ranked scores · Describes the strength and direction of the relationship · Uses ‘at least’ ordinal data Wilcoxon To test the significance of the difference. · A repeated measures design will have been used · Level of data is at least ordinal · The data has a normal distribution Mann Whitney-U test To test the significance of the difference. · An independent measures design will have been used · Level of data will be ‘at least’ ordinal