Research methods Revision Notes

What is a Hypothesis?

A hypothesis is a clear, testable, and falsifiable statement that predicts a relationship between variables. It's an educated guess or preliminary explanation that guides a research study. For example, 'Eating breakfast improves concentration in morning lessons'. This is a directional hypothesis because it predicts the nature of the relationship (improvement). A non-directional hypothesis would state 'Eating breakfast affects concentration in morning lessons', without specifying the direction of the effect. Every research study starts with forming a hypothesis based on observation or existing theories. This allows researchers to systematically investigate their predictions.

Independent and Dependent Variables

The independent variable (IV) is the factor that a researcher manipulates or changes. It is the 'cause' in a cause-and-effect relationship. For instance, if a researcher studies the effect of sleep deprivation on memory, sleep deprivation is the IV. The dependent variable (DV) is the factor that is measured. It is the 'effect' thought to be influenced by the IV. In the sleep deprivation example, memory performance would be the DV. It's essential to clearly define and operationalise both variables so they can be measured objectively. Operationalisation means exactly how the variables will be measured or manipulated. For example, 'sleep deprivation' could be operationalised as 'less than 4 hours of sleep in 24 hours', and 'memory performance' as 'the number of words correctly recalled from a list of 20'.

Understanding Null and Alternative Hypotheses

In research, two main types of hypotheses are used. The alternative hypothesis (H1) is the researcher's prediction about the outcome of the study, suggesting a significant relationship or difference between variables. For example, 'There will be a significant difference in happiness levels between participants who engage in daily meditation and those who do not'. The null hypothesis (H0) is a statement of no effect, no difference, or no relationship between the variables being tested. For instance, 'There will be no significant difference in happiness levels between participants who engage in daily meditation and those who do not'. The goal of statistical analysis in research is often to determine if there is enough evidence to reject the null hypothesis in favour of the alternative hypothesis.

Operationalising Variables for Measurement

Operationalisation is the process of making abstract concepts or variables measurable. It involves defining exactly how each variable will be observed or measured in a specific research study. For example, if a researcher wants to study 'stress', they might operationalise it as 'the score on a recognised stress questionnaire', 'heart rate', or 'number of reported anxious thoughts in a day'. Without clear operational definitions, it would be difficult to replicate studies, compare results across different studies, or determine the validity of the findings. It ensures that variables are concrete and quantifiable, moving psychological research from abstract ideas to empirical investigation.

Types of Data: Qualitative vs. Quantitative

Psychological research relies on two main types of data. Qualitative data is non-numerical and descriptive, providing rich detail and insight into experiences, feelings, and behaviours. Examples include interview transcripts, observational notes, or open-ended survey responses. It often deals with 'why' and 'how' questions. Quantitative data is numerical and measurable, allowing for statistical analysis. Examples include scores on a personality test, reaction times, number of errors made, or ratings on a scale. It focuses on 'what' and 'how much'. Choosing the right type of data depends on the research question and the depth of understanding required.

Measures of Central Tendency: Mean, Median, and Mode

Descriptive statistics summarise quantitative data. Measures of central tendency identify the 'average' or 'typical' value in a dataset. The Mean is calculated by summing all values and dividing by the number of values; it's sensitive to outliers. The Median is the middle value when data is arranged in order; it's less affected by extreme values. The Mode is the most frequently occurring value; it's useful for categorical data and can have multiple modes. Each measure provides a different perspective on the data's centre, and the choice depends on the data type and distribution.

Measures of Dispersion: Range and Standard Deviation

Measures of dispersion describe the spread or variability of data, indicating how scattered the scores are. The Range is the difference between the highest and lowest values; it's simple but can be heavily influenced by outliers. The Standard Deviation is a more sophisticated measure that quantifies the average amount by which scores deviate from the mean. A small standard deviation indicates scores are clustered closely around the mean, while a large standard deviation suggests scores are widely spread. It's often preferred for its precision and use in further statistical analysis.

Applying Descriptive Statistics

To effectively use descriptive statistics, you must first identify the type of data collected. For quantitative data, calculate appropriate measures of central tendency and dispersion. For example, if examining test scores, calculate the mean to see the average performance and the standard deviation to understand score variability. If comparing two groups, analyse their respective descriptive statistics to identify differences in typical performance and spread. If the data contains extreme scores (outliers), the median might be a more representative measure of central tendency than the mean. Always consider the context of the data and the research question when selecting and interpreting descriptive statistics.

What is Bias?

Bias in psychological research is a systematic error that can lead to inaccurate conclusions. It's not about making a mistake by chance, but rather a consistent deviation from the true results. Bias can affect the internal validity (whether the study accurately measures what it intends to) and external validity (whether the results can be generalized to other populations or settings) of a study. Identifying potential sources of bias is a critical step in conducting ethical and scientifically sound research.

Sampling Bias

Sampling bias occurs when the sample of participants in a study is not truly representative of the target population. This can happen if certain groups are over-represented or under-represented. For example, if a study on stress only recruits university students, the findings may not apply to older adults or individuals with different lifestyles. A common cause is convenience sampling, where researchers use easily accessible participants, who might share certain characteristics. Random sampling is a key method to reduce sampling bias, as it gives every member of the target population an equal chance of being selected.

Researcher Bias

Researcher bias, also known as experimenter bias, happens when the researcher's expectations, beliefs, or actions unintentionally influence the study's outcome or the participants' behavior. This can manifest in several ways, such as leading questions, differential treatment of participants, or selective interpretation of data. For instance, a researcher hoping to prove a hypothesis might unconsciously encourage participants to respond in a certain way. Double-blind procedures, where neither the participants nor the researchers know who is in which experimental group, are extremely effective in combating researcher bias.

Participant Bias

Participant bias, or subject bias, refers to the ways in which participants in a study may alter their behavior because they know they are being observed or because they have their own expectations about the study. The 'demand characteristics' phenomenon is a key example, where participants try to guess the aim of the study and behave in a way they think is expected of them. Another example is the 'social desirability bias,' where participants respond in a way that makes them look good or more socially acceptable, rather than giving their true feelings or behaviors. Using single-blind procedures, where participants do not know which group they are in, can help reduce participant bias.

Types of Data Displays

Psychological data can be presented in various ways to highlight different aspects. Tables are excellent for precise numerical values, often summarising raw data or descriptive statistics like means and standard deviations. Bar charts are typically used for comparing discrete categories or groups; the height of each bar represents the frequency or value for that category. Histograms are similar to bar charts but are used for continuous data, where the bars touch to show a distribution across a range of values. Pie charts show proportions of a whole, useful for illustrating how a total is divided into components. Line graphs display relationships between two continuous variables, often showing changes over time or across different conditions, with points connected to indicate trends.

Key Components of Graphs and Charts

All effective graphs and charts share common components that ensure clarity and accuracy. A clear title is essential, summarising the data being presented. Both the x-axis (horizontal) and y-axis (vertical) must be labelled clearly with the variable they represent and the units of measurement. Scales on the axes should be appropriate, starting from zero where possible to avoid exaggerating differences, and having sensible increments. A legend or key might be needed if multiple data sets or categories are plotted on the same graph, using different colours or symbols. For example, a line graph comparing anxiety levels of two different therapy groups would need a key to distinguish between the lines.

Interpreting and Evaluating Data Displays

Interpreting a table, chart, or graph involves extracting the key information and patterns. Look for trends, differences, and significant points. What is the highest value? The lowest? Are there any obvious correlations or changes over time? Evaluating these displays requires a critical eye. Consider if the chosen format is appropriate for the data type. Is the scale misleading? A truncated y-axis (one that doesn't start at zero) can make small differences appear much larger. Are the labels clear and unambiguous? Could the graph be misinterpreted? A good evaluation considers not just what the data shows, but also how effectively and fairly it is presented.