PRP1001/JHX1003 Research Methods 1 - Week 2 Lecture Notes PDF

**PRP1001/ JHX 1003** --------------------- **Class Notes** --------------- **Week 2** ---------- **The Open Science Movement** ----------------------------- The **reproducibility and replicability crisis** in scientific research, which highlighted the difficulty of consistently reproducing or r...

**PRP1001/ JHX 1003** --------------------- **Class Notes** --------------- **Week 2** ---------- **The Open Science Movement** ----------------------------- The **reproducibility and replicability crisis** in scientific research, which highlighted the difficulty of consistently reproducing or replicating the results of many studies has played a key role in driving the **open science movement**. Here\'s how the crisis and the movement are connected: - **Reproducibility** refers to the ability to **reproduce** the results of a study by reanalysing the original data using the same methods. - **Replicability** refers to the ability to **replicate** the findings by conducting a new study, following the same procedures, and getting the same results. Key reasons for the crisis include: - **Selective reporting** (or \"p-hacking\"), where only positive results are published. - **Small sample sizes** and weak experimental designs. - **Publication bias**, favouring novel and exciting findings over null or negative results. - **Lack of transparency** in data, methods, and materials. In response to these issues, the **open science movement** emerged, advocating for **transparency, accessibility, and collaboration** in scientific research. The goal of open science is to make research more **reliable, reproducible, and trustworthy** by promoting the sharing of data, methods, and findings. Here are the main ways the reproducibility and replicability crisis contributed to the movement: #### a. **Data Transparency and Open Data** - One major cause of the crisis was that researchers often **did not share their raw data**, making it difficult for others to reproduce the results. In response, the open science movement emphasises **open data**, where researchers are encouraged to share their datasets publicly so others can reanalyse the data and verify the results. - Open data allows for **reproducibility** because it provides the means for independent researchers to check if the original conclusions hold up under closer scrutiny. #### b. **Pre-registration of Studies** - The crisis revealed that researchers often engaged in practices like **p-hacking** (adjusting statistical analyses to make results appear significant) or **cherry-picking** data. To address this, the open science movement promotes **pre-registration** of studies, where researchers publicly declare their hypotheses, methods, and analysis plans **before** conducting their experiments. This practice reduces **confirmation bias** and increases the **transparency** of the research process. - Pre-registration ensures that what was planned is what is reported, helping to mitigate practices that can inflate false-positive findings. The crisis has also led to a shift in **scientific culture**, moving away from valuing only **novel findings** and toward valuing **transparency, replication, and reliability**. **Variables** ------------- **Variables** are the key elements in scientific research that represent measurable attributes or characteristics which can vary or change. They are used to test hypotheses and understand relationships between different factors in an experiment or study. In simpler terms, variables are anything that can be measured or manipulated in a scientific study. **Types of Variables** ---------------------- 1. **Independent Variable (IV)**: - The **independent variable** is the one that is **manipulated or changed** by the researcher to observe its effect. It is the \"cause\" in a cause-and-effect relationship. - Example: In an experiment to test whether sunlight affects plant growth, the **amount of sunlight** (hours of exposure) would be the independent variable. 2. **Dependent Variable (DV)**: - The **dependent variable** is the one that is **measured** or observed to see how it responds to changes in the independent variable. It is the \"effect\" in the cause-and-effect relationship. - Example: In the same plant experiment, the **growth of the plant** (height or number of leaves) would be the dependent variable, as it depends on the amount of sunlight. 3. **Controlled Variables (Constants)**: - **Controlled variables** are factors that are kept constant throughout the experiment to ensure that the results are due to changes in the independent variable, not other factors. - Example: In the plant experiment, variables like **soil type, water, and temperature** should be controlled to ensure that only the effect of sunlight is being tested. 4. **Confounding Variables**: - **Confounding variables** are unwanted variables that could influence the outcome of the experiment and **interfere** with the relationship between the independent and dependent variables. Researchers try to control or account for these to prevent biased results. - Example: If plants in the experiment receive different amounts of water (unintentionally), this could confound the relationship between sunlight and growth. **Examples** ------------ Below, you will find examples of study questions across three fields: **Psychology, Linguistics**, and **Sport Science**. Each example outlines the independent and dependent variables, measurement methods, and highlights potential control, confounding, and extraneous variables. A variety of methodologies are used to illustrate different approaches within each field. Before reading the explanations, start by just reading the study questions and try to identify the independent and dependent variables on your own. Then, think about potential control and confounding variables that could impact the results. Challenge yourself by reading the examples not only in your own field but in the other fields as well. ### **Psychology** ### Does transcranial magnetic stimulation (TMS) reduce symptoms of depression in patients with major depressive disorder (MDD)? - **Independent Variable**: TMS treatment (real TMS vs. sham TMS). - **Dependent Variable**: Depression symptom severity. - **Measurement**: Depression symptom severity will be measured using a standardised **questionnaire** (e.g., Beck Depression Inventory or BDI) before and after treatment. **Control Variables**: Medication use, age, and previous therapy involvement.\ **Confounding Variables**: Placebo effect due to participant expectations.\ **Extraneous Variables**: Sleep quality, diet, or physical activity levels. ### How do people with social anxiety process emotional expressions differently in terms of brain activation? - **Independent Variable**: Type of facial expression (fearful vs. neutral). - **Dependent Variable**: Brain activation patterns in response to emotional expressions. - **Measurement**: Participants will view a series of emotional faces while undergoing **fMRI** to measure brain activity in relevant regions, such as the amygdala. **Control Variables**: Presentation order of faces, participant age, and facial gender.\ **Confounding Variables**: Baseline anxiety levels may affect responses to emotional expressions.\ **Extraneous Variables**: Sleep patterns or physical health could influence brain activation. ### How does stress affect visual attention in individuals with high vs. low trait anxiety? - **Independent Variable**: Stress condition (stress-inducing task vs. non-stressful task). - **Dependent Variable**: Visual attention patterns. - **Measurement**: Attention will be assessed using **eye-tracking**, which measures where participants focus their gaze during a visual task. **Control Variables**: Task difficulty, screen brightness, and emotional content of stimuli.\ **Confounding Variables**: Coping strategies might influence how individuals react to stress.\ **Extraneous Variables**: Recent stressful life events or sleep patterns. ### Does music therapy reduce anxiety in patients with generalised anxiety disorder? - **Independent Variable**: Type of intervention (music therapy vs. no therapy). - **Dependent Variable**: Anxiety levels. - **Measurement**: Anxiety will be measured through **self-reported questionnaires** (e.g., State-Trait Anxiety Inventory) and **physiological measures** like heart rate and skin conductance. **Control Variables**: Age, medication use, and therapy duration.\ **Confounding Variables**: Participant preference for music could affect results.\ **Extraneous Variables**: Physical health and sleep quality. ### **Linguistics** ### How does word frequency affect reading comprehension and gaze duration in native and non-native English speakers? - **Independent Variable**: Word frequency (high-frequency words vs. low-frequency words). - **Dependent Variable**: Gaze duration on words during reading. - **Measurement**: **Eye-tracking** will be used to record gaze duration and fixation points as participants read a passage containing both high- and low-frequency words. **Control Variables**: Passage length, sentence structure, and participant proficiency level.\ **Confounding Variables**: Reading proficiency differences between native and non-native speakers, unrelated to word frequency.\ **Extraneous Variables**: Fatigue, familiarity with the topic of the passage, or room lighting could affect reading behaviour. ### How does sentence complexity affect brain responses to syntactic violations in native speakers? - **Independent Variable**: Sentence complexity (simple vs. complex syntax). - **Dependent Variable**: Brain response to syntactic violations. - **Measurement**: **EEG** will be used to record event-related potentials (ERPs), such as the **P600**, which is associated with syntactic processing when participants encounter grammatical violations in both simple and complex sentences. **Control Variables**: Sentence length, word frequency, and presentation speed.\ **Confounding Variables**: Differences in working memory capacity, which could affect processing of complex sentences more than simple ones.\ **Extraneous Variables**: External distractions, recent cognitive load, or general language exposure before the task. ### Do bilingual children develop executive control skills faster than monolingual children? - **Independent Variable**: Language background (bilingual vs. monolingual). - **Dependent Variable**: Executive control skills, such as response inhibition or cognitive flexibility. - **Measurement**: Executive control will be assessed through behavioural tasks (e.g., **Stroop task** or **Dimensional Change Card Sort task**) and **EEG** recordings to measure brain activity during these tasks. **Control Variables**: Age, socioeconomic status, and overall cognitive development level.\ **Confounding Variables**: Exposure to multiple languages or multilingual environments in monolingual children, which might influence executive control.\ **Extraneous Variables**: Sleep quality, recent stress, or task familiarity. ### How does visual complexity of written language affect reading strategies in individuals learning a logographic writing system? - **Independent Variable**: Visual complexity of characters (high-complexity logographic characters vs. low-complexity characters). - **Dependent Variable**: Reading strategies, such as fixation duration and number of regressions (re-reading parts of text). - **Measurement**: **Eye-tracking** will be used to measure gaze duration, saccades, and regressions as participants read texts with varying levels of character complexity. **Control Variables**: Sentence length, word frequency, and participant reading proficiency.\ **Confounding Variables**: Previous exposure to other complex writing systems, which could make participants more adept at processing visually complex characters.\ **Extraneous Variables**: Participants' motivation, reading fatigue, or recent experiences with the target language. ### **Sport Science** ### How does resistance training intensity affect muscle activation in professional athletes? - **Independent Variable**: Training intensity (low vs. high resistance). - **Dependent Variable**: Muscle activation levels. - **Measurement**: Muscle activation will be measured using **EMG** to record electrical activity in the muscles during resistance exercises. **Control Variables**: Type of exercise, participant fitness level, and time of day.\ **Confounding Variables**: Previous experience with resistance training could lead to differing adaptation rates between participants.\ **Extraneous Variables**: Fatigue, hydration levels, or recent injury could affect muscle activation independently of training intensity. ### How does aerobic endurance training affect cardiovascular adaptations in older adults? - **Independent Variable**: Type of training program (aerobic endurance training vs. no training). - **Dependent Variable**: Cardiovascular adaptations, including heart rate variability and VO2 max. - **Measurement**: Cardiovascular performance will be assessed by measuring **VO2 max** using a treadmill test, and **heart rate variability** using wearable heart monitors. **Control Variables**: Diet, medication use, and training duration.\ **Confounding Variables**: Pre-existing cardiovascular health conditions might influence adaptations to the training program.\ **Extraneous Variables**: Stress, environmental factors like air quality, and hydration status could impact cardiovascular measurements. ### How does sleep deprivation affect athletic performance and decision-making in basketball players? - **Independent Variable**: Sleep condition (24 hours of sleep deprivation vs. full night's sleep). - **Dependent Variable**: Athletic performance and decision-making accuracy. - **Measurement**: **Physical performance** will be assessed using basketball drills (e.g., shooting accuracy, sprinting speed), and decision-making will be measured through **reaction time tasks** and in-game decision-making scenarios. **Control Variables**: Nutrition, pre-exercise warm-up routines, and time of testing.\ **Confounding Variables**: Natural sleep patterns or individual resilience to sleep deprivation might impact results.\ **Extraneous Variables**: Caffeine intake, stress levels, or team dynamics during game-like scenarios. ### How does hydration level affect muscle fatigue during prolonged exercise in endurance athletes? - **Independent Variable**: Hydration condition (well-hydrated vs. dehydrated). - **Dependent Variable**: Muscle fatigue and endurance performance. - **Measurement**: Muscle fatigue will be measured using **EMG** to assess electrical activity during sustained exercise, and endurance performance will be assessed using time-to-exhaustion tests. **Control Variables**: Environmental conditions, exercise intensity, and nutrition intake.\ **Confounding Variables**: Natural sweat rates or differences in hydration strategies could influence muscle fatigue levels.\ **Extraneous Variables**: Mental fatigue, stress, or external distractions during the performance test. ### How does strength training affect neural adaptations in novice weightlifters? - **Independent Variable**: Strength training program (strength training vs. no training). - **Dependent Variable**: Neural adaptations, specifically motor cortex activity. - **Measurement**: **fMRI** will be used to assess brain activity related to motor control during specific lifting tasks before and after the training intervention. **Control Variables**: Type of strength training exercise, training frequency, and sleep patterns.\ **Confounding Variables**: Previous athletic experience or genetic predisposition to faster neural adaptations.\ **Extraneous Variables**: Stress, caffeine intake, or distractions during the fMRI sessions. ### **Formulating Research Questions** Good research questions often have clear **independent** and **dependent** variables, as this helps establish a clear and testable relationship between cause and effect. A well-defined research question provides a solid foundation for designing an experiment or study, making it easier to identify what you will manipulate (independent variable) and what you will measure (dependent variable). ### Example of a Good Research Question: - Here, the question has a clear independent variable (**amount of daily exercise**) and a clear dependent variable (**memory retention**), making it a strong, focused research question. In correlational designs, the relationship between variables is studied without the researcher manipulating any variables, so technically, there isn\'t a true independent or dependent variable as there would be in an experimental design. However, it is still useful to identify the variables involved and clarify which one is thought to influence or be associated with the other. ### Example of a Good Research Question in a Correlational Study: In this case, while no variable is directly manipulated, we can still think of hours of sleep as the predictor variable (what might influence) and levels of stress as the outcome variable (what is influenced). ### **Operationalising Variables** Operationalisation is the process of turning variables into something that can be empirically observed and tested. Operationalising variables means defining how abstract concepts or variables will be measured and manipulated in a specific, concrete way for the purpose of a research study. It transforms theoretical concepts (which may be broad or abstract) into measurable observations or data points. Operationalisation applies to both independent and dependent variables, as well as any other key variables in a study. The process of operationalisation is used to clearly define how any variable will be measured or manipulated in a specific and observable way, regardless of whether it is an independent, dependent, or control variable. Example: **\"Does increased motivation improve academic performance in college students?\"** - **Independent Variable**: **Motivation** - This could be operationalised using a **self-report survey**, measuring motivation on a scale (e.g., Academic Motivation Scale), or by using **behavioural indicators** such as study hours or class attendance. - **Dependent Variable**: **Academic performance** - This could be operationalised as **GPA**, final exam grades, or the number of assignments submitted on time. Both variables need to be operationalised to ensure that they can be measured accurately and consistently. When operationalising variables, there can be many different definitions for a single hypothetical construct. The construct of **intelligence** is an example with multiple operational definitions: 1. **IQ Tests**: Intelligence could be measured by a standardised test, such as the **Wechsler Adult Intelligence Scale (WAIS)**, which provides an **IQ score**. 2. **Problem-Solving Tasks**: Intelligence could be operationalised as performance on **problem-solving tasks**, where speed and accuracy are measured. 3. **Creative Thinking**: Intelligence could be measured by assessing **creative problem-solving** or **divergent thinking**, which reflects a different aspect of intelligence. 4. **Academic Achievement**: In some contexts, intelligence might be operationalised by **grades** or **test scores** in school subjects. **Reification** **Reification** is the process of treating an abstract concept or hypothetical construct as if it were a concrete, tangible thing. It involves giving **physical reality** to something that is not inherently physical or directly measurable, which can lead to misunderstanding or misrepresentation of the concept. In research, this can be problematic because it makes abstract ideas seem more definite or \"real\" than they truly are, potentially oversimplifying complex psychological or social phenomena. ### **Example: Motivation** **Motivation** is an abstract construct---an internal drive that influences behaviour, which is not directly observable. It can be operationalised and measured through various means (*e.g*., self-reports, observed behaviour, physiological indicators), but it is not something that exists as a physical object or entity. If researchers treat **motivation** as a **concrete thing** rather than an abstract concept, they might assume it works the same way in all contexts, oversimplifying its nature. This could lead to the mistaken belief that motivation is a fixed or quantifiable \"thing\" that exists in a person, rather than a complex, multifaceted process influenced by various factors. ### **Issues with Reification in Different Fields** #### **1. Psychology:** In **psychology**, reification of motivation might result in treating it as a **fixed trait** rather than a **dynamic process** that fluctuates based on context, mood, and environmental factors. For instance, if a psychologist reifies motivation, they might overgeneralise and say, \"This person has low motivation,\" as though it were a constant trait, rather than considering that their motivation could vary depending on the task, time, or setting. - **Problem**: This could lead to **inflexible interventions**, where treatments or strategies are based on the assumption that motivation is static, rather than focusing on ways to increase or manage motivation in different situations. #### **2. Sport Science:** In **sport science**, motivation is crucial for understanding **athlete performance**. Reifying motivation might lead to thinking of it as a **quantity** that an athlete either possesses or lacks. For example, if motivation is reified, a coach might say, \"This athlete is unmotivated,\" as if motivation were an inherent, unchangeable trait. - **Problem**: This perspective could result in ignoring **external factors** (e.g., stress, fatigue, or competition environment) that may influence an athlete's motivation. Reification may lead to simplistic conclusions and ineffective training strategies because the complex and changing nature of motivation is not acknowledged. #### **3. Linguistics:** In **linguistics**, motivation can also relate to how learners are driven to acquire a second language. Reifying motivation in this context could lead to assuming that students either \"have motivation\" or \"lack motivation\" to learn a language, without considering the various **types of motivation** (e.g., **intrinsic** vs. **extrinsic**) or how motivation may evolve throughout the learning process. - **Problem**: This might result in ineffective teaching methods, as educators might assume that motivation is a fixed attribute in students, failing to adapt teaching approaches to **foster or enhance motivation** dynamically in the language-learning environment. **Scales of measurement** Stanley Stevens\' **(1946)** theory of **scales of measurement** is a classification system that categorises different ways variables can be measured in research. He introduced four levels of measurement, each with increasing complexity and precision. These scales help researchers determine which types of **statistical analyses** are appropriate for the data they collect. The four scales are **nominal, ordinal, interval, and ratio**. ### 1. **Nominal Scale**: - **Definition**: The nominal scale is the simplest level of measurement. It involves **categorising** data into distinct groups or categories that have **no numerical or quantitative value**. - **Characteristics**: Categories are **mutually exclusive** and **unordered**. There is no ranking or hierarchy. - **Examples**: - Gender (male, female, non-binary) - Types of pets (dog, cat, bird) - Colours (red, blue, green) ### ### 2. **Ordinal Scale**: - **Definition**: The ordinal scale involves categories that can be **ranked or ordered** in a meaningful way, but the **differences between the ranks are not equal or known**. - **Characteristics**: Data can be ranked, but the intervals between the ranks are not necessarily consistent or measurable. - **Examples**: - Education level (high school, undergraduate, graduate) - Satisfaction ratings (very dissatisfied, dissatisfied, neutral, satisfied, very satisfied) - Race placement (1st place, 2nd place, 3rd place) ### 3. **Interval Scale**: - **Definition**: The interval scale is a more sophisticated scale in which the data can be ordered, and the **intervals between values are meaningful and equal**, but there is **no true zero point**. - **Characteristics**: Data can be ranked, and the difference between values is meaningful, but ratios (e.g., \"twice as much\") cannot be calculated because there is no absolute zero. - **Examples**: - Temperature in Celsius or Fahrenheit (the difference between 20°C and 30°C is meaningful, but there is no \"true zero\" of temperature) - IQ scores - Calendar years (2000, 2020, etc.) ### 4. **Ratio Scale**: - **Definition**: The ratio scale has all the properties of the interval scale, but with an **absolute zero** point, meaning that zero represents a complete absence of the quantity being measured. This allows for meaningful **ratios** between values. - **Characteristics**: Data can be ordered, intervals are equal, and meaningful ratios can be calculated (e.g., \"twice as much\"). Zero means the absence of the variable. - **Examples**: - Temperature in Kelvin. Kelvin has an absolute zero (0 K), which represents the complete absence of thermal energy. This absolute zero makes ratios meaningful on the Kelvin scale. For example, 200 K is twice as hot as 100 K. - Weight (a weight of 0 means no weight) - Height (a height of 0 means no height) - Reaction time (a reaction time of 0 would mean no reaction) - Distance (zero means no distance travelled) **Why Is a True Zero Important?** A true zero allows for ratio comparisons and gives a clear meaning to the absence of the variable being measured. Without a true zero, scales are limited to adding or subtracting values, but you can\'t make proportional comparisons like \"twice as much\" or \"half as much.\" This is why ratio scales (with a true zero) are more flexible and allow for more mathematical operations compared to interval scales, which lack a true zero. **Accuracy of Measurement** **Accuracy of measurement** is crucial in research because it ensures that the data collected is reliable, precise, and meaningful. The more accurate the measurement, the better the ability to draw valid conclusions from the research and apply appropriate statistical analyses. Would it be better to measure height using **\"tall, medium, short\"** or in **centimetres**? The answer depends on the **accuracy** and **detail** required for the research. 1. **\"Tall, Medium, Short\"**: - This is an **ordinal scale** of measurement, where height is categorised into ranked groups without precise intervals. While this may be useful for **simple categorisation** or when exact measurements are not necessary, it lacks precision. The categories may be **subjective** and can vary depending on cultural or individual perceptions of what constitutes \"tall\" or \"short.\" This limits the ability to perform detailed statistical analyses and compare heights accurately. 2. **Centimetres**: - Measuring height in **centimeters** provides a **ratio scale**, which offers far more precision and accuracy. This scale allows for meaningful comparisons (*e.g*., one person being exactly 5 cm taller than another) and also supports more advanced **statistical analysis** (e.g., calculating means, variances, correlations). Additionally, centimetres have a **true zero** point (0 cm represents no height), making this measurement method more detailed and reliable. ### **Correlational Design** A **correlational design** is a type of non-experimental research that examines the **relationship** or **association** between two or more variables without manipulating them. Unlike experimental designs, where variables are controlled and manipulated, correlational studies simply **measure** variables to see if they are related. The strength and direction of the relationship are expressed using a **correlation coefficient** (*e.g*., Pearson's r), which ranges from -1 to +1. A value of **+1** indicates a perfect positive correlation, **-1** indicates a perfect negative correlation, and **0** means no relationship. ### **Examples of Correlational Design in Different Fields** #### **1. Psychology Example:** **Research Question**: Is there a relationship between **stress** and **sleep quality** in college students? In this example, a psychologist might measure the **stress levels** (using a stress questionnaire) and **sleep quality** (using a self-report scale or sleep monitoring device) of students. By analysing the data, the researcher could determine if there is a **positive** or **negative** relationship between these two variables. A **negative correlation** might show that as stress increases, sleep quality decreases. However, this does not mean that stress **causes** poor sleep; there could be other factors at play, such as lifestyle or health conditions. **Interpretation**: A correlational study like this helps identify whether two variables are related but cannot establish why they are related. #### 2. **Sport Science Example:** **Research Question**: Is there a relationship between **athletes' fitness levels** and their **injury rates**? A sport scientist could measure **fitness levels** (e.g., using VO2 max or strength assessments) and track **injury rates** (number of injuries in a season) in a group of athletes. By conducting a correlational analysis, the researcher can determine if there is an association between higher fitness levels and lower injury rates, or vice versa. This information can guide further exploration into **why** these variables are related, but the correlational design itself would not establish causality. **Interpretation**: The researcher might find that better fitness correlates with fewer injuries, but other factors, such as **training methods** or **nutrition**, could also be involved. #### 3. **Linguistics Example:** **Research Question**: Is there a relationship between **exposure to second-language media** and **language proficiency** in second-language learners? In this case, a linguist could survey participants on how much they engage with media in their second language (e.g., watching movies, listening to music) and assess their **language proficiency** using a standardised test. A correlational analysis would reveal whether higher exposure to media in the second language is associated with higher language proficiency. **Interpretation**: While this study might find a positive correlation between media exposure and language proficiency, it does not establish that media exposure **causes** better proficiency. There may be other factors, like **formal language education** or **prior experience**, influencing the results. ### **Experimental Design** An **experimental design** is a type of research method where the researcher **manipulates one or more independent variables** to observe their effect on a **dependent variable**. This type of design allows researchers to establish **cause-and-effect relationships** between variables. Participants are typically assigned to different groups, such as **treatment** and **control** groups, and the researcher measures the outcomes after introducing a specific intervention or manipulation. **Random assignment** is often used to ensure that any observed effects are due to the manipulation and not to pre-existing differences between participants. ### **Examples of Experimental Design in Different Fields** #### **1. Psychology Example:** **Research Question**: Does **daily exercise** reduce **stress levels** in college students? In this study, participants are randomly assigned to one of two groups: an **exercise group** that performs 30 minutes of daily exercise, and a **control group** that does not perform any additional exercise. After four weeks, **stress levels** are measured using a standardised self-report stress scale. **Interpretation**: If the exercise group shows lower stress levels than the control group, this suggests that daily exercise can lead to a reduction in stress. The experimental design allows for the possibility of establishing a **causal relationship** between exercise and reduced stress levels. #### 2. **Sport Science Example:** **Research Question**: Does practicing **yoga** improve **flexibility** in amateur athletes? In this experiment, athletes are randomly assigned to either a **yoga group**, which practises yoga for 30 minutes a day, or a **no-intervention group**. After six weeks, the **flexibility** of each participant is measured using a **range of motion test** for specific joints (e.g., hamstrings, hips). **Interpretation**: If the yoga group shows greater improvements in flexibility compared to the no-intervention group, the results suggest that practising yoga causes an increase in flexibility. The experimental design enables the researcher to make causal inferences about the effect of yoga on flexibility. #### 3. **Linguistics Example:** **Research Question**: Does **speaking practice** improve **pronunciation accuracy** in language learners? Language learners are randomly assigned to an experimental group that practises speaking for 30 minutes daily and a control group that does not engage in additional speaking practice. After eight weeks, **pronunciation accuracy** is assessed using a pronunciation test where native speakers rate the accuracy of participants\' speech. **Interpretation**: If the speaking practice group demonstrates improved pronunciation accuracy compared to the control group, it suggests that regular speaking practice leads to better pronunciation. The experimental design supports a **causal conclusion** about the impact of speaking practice on pronunciation accuracy. In experimental design, you can use either a **between-subjects design** or a **within-subjects design**, depending on how you assign participants to groups and how you measure the variables. Both designs are valuable, but they serve different purposes and have different advantages. ### **Between-Subjects Design**: In a **between-subjects design**, participants are **randomly assigned to different groups**, and each group experiences a different condition of the independent variable. Each participant is only exposed to **one experimental condition**, so comparisons are made **between** the groups. #### Example: Stress and Sleep Quality (Between-Subjects Design) **Research Question**: Does reducing stress improve sleep quality in college students? - **Design**: Participants are randomly assigned to one of two groups: - **Experimental group**: Receives a stress-reduction intervention (e.g., relaxation techniques). - **Control group**: Does not receive any intervention. - **Outcome**: After several weeks, **sleep quality** is measured for both groups. - **Analysis**: You would compare the sleep quality between the two groups to see if the intervention group showed greater improvements than the control group. **Advantage**: No **carryover effects** (effects from one condition influencing the next), as participants only experience one condition. ### **Within-Subjects Design**: In a **within-subjects design**, the **same participants** are exposed to **all experimental conditions**. Each participant experiences every condition of the independent variable, and their performance or responses are compared across those conditions. #### Example: Stress and Sleep Quality (Within-Subjects Design) **Research Question**: Does reducing stress improve sleep quality in college students? - **Design**: All participants undergo both conditions: 1. **Baseline period**: No stress-reduction techniques are used. 2. **Intervention period**: Participants practise stress-reduction techniques. - **Outcome**: **Sleep quality** is measured during both periods for each participant. - **Analysis**: You would compare each participant's sleep quality during the baseline period and the intervention period. **Advantage**: Reduces **individual differences** between groups since each participant serves as their own control. This design typically requires **fewer participants**.