Practical Investigation (Edexcel A-Level Psychology): Revision Notes
Guide to the Practical Investigation
Learning outcomes
This practical investigation will equip you with essential research skills for analysing media representations of mental health issues. By the end, you'll be able to conduct systematic content analysis and understand the ethical implications of such research.
This practical investigation will teach you how to:
- Conduct a summative content analysis on a clinical psychology topic
- Gather data from two sources relevant to attitudes towards mental health
- Perform a summative analysis using data from two sources
- Explore ethical issues associated with this type of research
What is summative content analysis?
Content analysis is a research technique that transforms qualitative data into quantitative data. Different types of content analysis have developed to serve various research purposes.
Understanding Content Analysis
A content analysis is a technique used to quantify qualitative data. This means taking text, images, or other media and converting observations into numerical data that can be analysed statistically.
Types of content analysis
Conventional content analysis begins with the raw data itself. Researchers review and analyse the material, allowing coding categories to emerge naturally from what they observe in the data.
Directive content analysis starts with predetermined coding categories. These categories come from existing theories or previous research findings. While researchers may modify these categories during analysis, the initial framework is established before examining the raw data. This approach is particularly useful when extending or testing existing knowledge about a topic.
Summative content analysis focuses on counting how often specific key words or terms appear in the data. These key words can be identified either before starting the analysis or during the process. The approach goes beyond simple counting—researchers must examine the context in which these words appear, including who said them, how they were used, and to whom they were directed.
Key Distinction: CDS Framework
Remember the three types using CDS:
- Conventional: Categories emerge from the data
- Directive: Categories derived from existing theory
- Summative: Counting and contextualising key words
The summative content analysis process
For this practical investigation, you will use summative content analysis with elements of conventional analysis. The process involves:
- Identifying key words of interest related to your topic
- Calculating the frequency of these key words in two different sources (manifest content)
- Moving beyond surface-level counting to interpret the meaning and context of these words (latent content)
For example, you might compare how different newspapers discuss the same topic by examining their use of key words related to mental health issues. In the context of clinical psychology, this practical will focus on analysing two films that depict mental illness.
Developing your research question
Your research question must be answerable using summative content analysis and must focus on attitudes towards mental health issues.
Example Research Question
"Have media depictions of mental health issues and treatment changed over the years?"
This type of question allows you to compare two sources from different time periods to identify changes in how mental health is portrayed. It's specific enough to guide your keyword selection while broad enough to allow for meaningful analysis.
Selecting your sample
The sample consists of the materials you will analyse. You need to select two sources that allow meaningful comparison. These sources should be of the same type (e.g., both films) but differ in a relevant way (e.g., era of production).
Source selection examples
A quick search reveals films such as:
- One Flew Over the Cuckoo's Nest (1975) and Girl, Interrupted (1999)—both address mental health issues and therapy from different eras
- Psycho (1960) and Fight Club (1999)—both depict dissociative identity disorder from different time periods
Practical considerations
Time Management Tip
Transcribing qualitative data into text form is time-consuming, even for small amounts of content. If you are using film sources, restrict your analysis to a five- or ten-minute clip to make the task manageable. This focused approach still provides valuable data while keeping the workload realistic.
Defining keywords
From your research question, formulate a hypothesis. For example: "There will be a greater number of negative references to people with mental disorders in an older film compared to a more recent film."
This hypothesis allows you to examine whether media depictions of mental disorders have changed over time.
Identifying keywords
You can determine key words before or during your review of the material. For this investigation, you are examining how mental illness is depicted in films, particularly focusing on derogatory terms used to describe people with mental illness.
Sample Keyword List for Mental Health Depictions
Examples of potential keywords include:
- Loony
- Nutcase
- Crazy
- Dangerous
- Has a screw loose
- Lost their marbles
- Mad
- Insane
- Cuckoo
- Psycho
- Schizo
- Loopy
Note: This list can be modified or expanded once you begin your review of the material. The keywords should reflect terms that indicate attitudes toward mental health.
Conducting the analysis
Manifest content analysis
Once you have transcribed the film material, review the content and count how often each key word appears. Compare these frequencies across both films. Remember that the predetermined key words can be modified or expanded once your review begins.
Understanding Manifest Content
Manifest content refers to the surface-level, visible elements of your data. In this case, it means the straightforward frequency count of how many times each keyword appears in your sources. This provides the quantitative foundation for your analysis.
Ensuring validity and reliability
Inter-coder reliability is essential for maintaining objectivity. An expert in clinical research should verify that your key terms are valid, and multiple coders should analyse the same material. If different coders show good agreement, inter-coder reliability is established.
Critical for Research Quality
Without inter-coder reliability, your findings may reflect personal bias rather than objective patterns in the data. Always have at least one other person independently code a portion of your material to verify consistency.
Tallying the data
Counting the frequency of data in specific categories produces nominal data. This data is best summarised using percentages or proportions of the total.
Latent content analysis
You must go beyond simply counting key word frequencies. Consider the context in which these words were used. In this practical, examine:
- The people involved in the dialogue
- The direction of communication
Understanding Latent Content
Latent content refers to the deeper meaning and context behind the surface-level data. While manifest content tells you "how many times," latent content helps you understand "what it means" and "why it matters."
For latent content analysis, determine:
- Whether the word was spoken by a patient to another patient
- Whether it was said by a patient to a health professional
- Whether a health professional used it to address a patient
- Whether a health professional used it when speaking to another health professional
- Other communication patterns
Presenting your data
Coding sheet example
Create a table to record the frequency of key words and their context. Table 5.5 shows an example structure:
| Communication Type | Crazy | Loony | Nutcase | Mad |
|---|---|---|---|---|
| Patient to patient | 111111 | 1 | 1111111 | 111 |
| Patient to mental health professional | 11 | 111111 | ||
| Mental health professional to mental health professional | 1111 | 1111111 | 1 | 11111 |
| Mental health professional to patient | 1 | 1111111111 | 111 | 11 |
| Patient to relative/friends | 11 | 11 | 1 | |
| Relative/friends to patient | 1 | 1 | 1 | |
| Other (specify relationship) | To taxi driver 1 | By taxi driver to patient 11 |
Recording Frequencies
Each tally mark (1) represents one occurrence of the keyword. This manual counting method helps you maintain awareness of patterns as they emerge, rather than relying solely on automated searches.
Converting to percentages
Transform the raw frequency data into percentages to show the proportions more clearly. For example:
| Communication Type | Crazy |
|---|---|
| Patient to patient | 36% |
| Patient to mental health professional | 12% |
| Mental health professional to mental health professional | 24% |
| Mental health professional to patient | 6% |
| Patient to relative/friends | 12% |
| Relative/friends to patient | 6% |
| Other (specify relationship) | 0% |
Why Use Percentages?
Converting raw frequencies to percentages allows for easier comparison between sources that may have different total word counts or lengths. A film with more dialogue would naturally have more keyword occurrences, but percentages standardise the comparison.
Visual representation
You can present the data graphically using charts. A pie chart effectively shows the proportions of different communication types, making patterns easier to identify visually.
Drawing conclusions
Summative content analysis is time-consuming but highly flexible across different materials. It provides an objective record of communication patterns that might not be immediately obvious, revealing deeper latent issues.
When interpreting the latent content, examine:
- Relationships between participants
- Trends across different sources
- Comparisons between the sources
- What these patterns suggest about the material
Support your interpretations with evidence from the data.
Sample Interpretation: Analysing the Word 'Crazy'
An analysis might reveal that the word 'crazy' appeared more frequently in patient-to-patient communication and between mental health professionals. This pattern suggests that stereotypical notions of mental illness were being transmitted not only casually between patients but also between health professions—indicating that stereotypes of mental illness were institutional terms of reference.
However, if the key word is not directed at the patient by a mental health professional, this suggests some level of professionalism or clinical guidelines regarding stereotypical language when addressing patients. Compare this interpretation with the second source to determine if the same frequency and communication patterns exist.
Key point: The pattern reveals not just the presence of stigmatising language, but where and how it circulates within the institutional setting.
Evaluation
Strengths
- Provides objective analysis of how key words are used in media
- Reveals patterns and context during the quantification process
- Can be applied flexibly across various materials
Weaknesses
- Extremely time-consuming to conduct properly
- The actual meaning and context of key words may be lost during quantification
- For example, a patient might say 'screw loose' to another patient as a euphemism describing how health professionals view them, not because the patient is mentally ill
- This demonstrates how meaningfulness can be lost when quantifying qualitative data
Common Pitfall: Loss of Nuance
When quantifying qualitative data, there's always a risk of losing subtle meanings and context. A word like "crazy" might be used ironically, affectionately, or clinically—but frequency counts alone don't capture these distinctions. This is why latent content analysis is essential to complement manifest analysis.
Considerations for improvement
When evaluating your practical investigation, consider:
- Generalisability of your sources
- Inter-coder reliability
- Validity of key terms and coding
- Credibility of findings
- Objectivity of the analysis
- Subjectivity in interpretation
- Possible changes to improve the investigation
Ethical considerations
Protection and permissions
Although you are not dealing with actual participants, you must ensure:
- The information analysed is not private or confidential
- The material is in the public domain, or you have explicit permission from those whose material you are analysing
- The material is not publicly available only if permission has been granted
Socially sensitive research
Handling Socially Sensitive Findings
Consider the potential impact of your findings on different groups in society. If your conclusions could be socially sensitive, think carefully about whether making them public might be detrimental to certain groups.
For example, if you found that certain cultural groups have more negative attitudes than others, publicising this finding could be harmful to those groups. The potential for stigmatisation must be weighed against the scientific value of the findings.
Avoiding bias and overstatement
Do not overstate or sensationalise your findings. Verify your interpretations with a more experienced researcher, as the data and categories represent your own subjective interpretation of the material being analysed. The coding categories and analysis reflect personal judgement, which introduces potential bias.
Practical issues in design and implementation
For reliable and objective data collection, clear operationalisation of coding is required. Developing coding units that are clear and unambiguous removes potential sources of bias, allowing other researchers to replicate your method and test the reliability of your conclusions.
Ensuring Replicability
Summative analysis uses a key word search initially, which provides objectivity in implementation. However, you must ensure that the key words being used accurately measure what you intend to record. Clear definitions and examples of each keyword help other researchers replicate your study and verify your findings.
Remember!
Key Points to Remember:
- Summative content analysis quantifies qualitative data by counting key words in context
- You must analyse both manifest content (frequencies) and latent content (meaning and context)
- Select two comparable sources that allow you to answer your research question about attitudes to mental health
- Ensure inter-coder reliability by having multiple people code the same material
- Present your findings clearly using tables and visual representations
- Consider ethical issues carefully, particularly when dealing with socially sensitive topics like mental health
- The SQKA process guides your work: Sample selection, Question development, Keywords definition, Analysis execution
- Evaluate your study using VROG: Validity, Reliability, Objectivity, Generalisability