Lesson 10 of 10 · 20 min
Researching the brain
Apply the science inquiry skills to brain research, from case studies and scans to a reaction-time experiment you could run, with variables, hypotheses, design, controls, ethics and descriptive statistics.
What you will be able to do
- Identify the type of investigation behind evidence about the brain, and what it can and cannot show.
- Identify and operationalise the independent and dependent variables of an experiment, and write null and alternative hypotheses.
- Explain controls such as counterbalancing and standardised procedures, and apply ethical principles.
- Calculate and interpret the mean, median, range and standard deviation of a data set.
Everything in this topic is the product of research, and the methods behind it differ a great deal in what they can show. Psychologists judge a claim by asking how the evidence was collected. This lesson uses the topic's own studies, and a simple experiment you could run in class, to practise the science inquiry skills that run through the whole course.
| Type of investigation | Example from this topic | What it can show | Main limitation |
|---|---|---|---|
| Case study | Broca's patient Leborgne; Phineas Gage | Detailed evidence that damage to an area goes with a change in ability; rare cases nothing else could study | One person: may not apply to others; damage not controlled; often written up long after |
| Experiment | Flourens' ablation; a reaction-time experiment | Cause and effect: the researcher manipulates the independent variable and controls other variables | Animal results may not apply to humans; ethics limit what can be done to people |
| Quasi-experiment | Raine (1997): people charged with murder (pleading insanity) and controls compared with PET | Differences between groups that already exist and cannot be assigned at random | Groups may differ in other ways too, so cause cannot be concluded |
| Correlational study | Hours of sleep and EEG activity | Whether two variables are related, and how strongly | A relationship does not show cause |
| Observational study | Recording the behaviour of patients on a stroke ward | Behaviour as it naturally happens | Observers can be biased; no control of variables |
Definition
Variables
In an experiment, the researcher changes the independent variable (IV) and measures its effect on the dependent variable (DV), keeping other variables the same (controlled variables). An extraneous variable is any other variable that could affect the DV; if it changes along with the IV, it is a confounding variable, and the result can no longer be put down to the IV.
Each variable must be operationalised: defined by exactly how it is manipulated or measured. "Reaction speed" is vague; "mean time in milliseconds to click when a square turns green, over 20 trials" is operationalised.
A class experiment. The left primary motor cortex controls the right hand, and most people use one hand far more than the other. Does that practice make the dominant hand react faster? A class designs an experiment:
- Eight Year 11 students each do an online reaction-time test: click as soon as a square on the screen turns green. Each student does 20 trials with the dominant hand and 20 with the non-dominant hand.
- Half the students start with the dominant hand and half with the non-dominant hand.
- The DV is each student's mean reaction time, in milliseconds, for each hand.
Classify each variable, choose the operational definitions of the IV and DV, then pick the hypotheses.
Key idea
Hypotheses
The null hypothesis (H₀) states that the IV has no effect: "There will be no difference between the mean reaction times of the dominant and non-dominant hands."
The alternative hypothesis (H₁) states that it does: "The mean reaction time of the dominant hand will be shorter than that of the non-dominant hand." Write it with the operationalised IV and DV and the population (here, Year 11 students).
In the reaction-time experiment, what is the independent variable?
Design and controls. Each student is tested in both conditions, so this is a repeated measures design. Its strength is that differences between people (some students are simply quicker) affect both conditions equally. Its weakness is order effects: students may improve with practice, or tire, by the second condition. Counterbalancing, half starting with each hand, spreads order effects evenly across both conditions. Standardised instructions and procedures (the same script, computer, chair and number of practice trials for everyone) remove variation in how students are treated.
Ethics. Even a harmless class experiment follows ethical principles: informed consent in writing (with written permission from a parent or carer for any student under 16, as QCAA's guidelines for school investigations require), voluntary participation and the right to withdraw at any time, confidentiality (results recorded by code number, not name), and a debriefing that explains the aim afterwards. The history of brain research is the reason these rules exist: Flourens' animals and many early patients were never protected by them.
| Student | Dominant hand | Non-dominant hand | Difference |
|---|---|---|---|
| 1 | 252 | 271 | 19 |
| 2 | 241 | 255 | 14 |
| 3 | 268 | 284 | 16 |
| 4 | 230 | 249 | 19 |
| 5 | 275 | 280 | 5 |
| 6 | 259 | 277 | 18 |
| 7 | 247 | 262 | 15 |
| 8 | 236 | 250 | 14 |
Worked example
Describe reaction-time data with the mean and standard deviation
The dominant-hand reaction times (ms) of the eight students in the class experiment were:
252, 241, 268, 230, 275, 259, 247, 236
Calculate (a) the mean, (b) the median, (c) the range and (d) the sample standard deviation, each to 2 decimal places.
Step 1: (a) Mean
Sum , so ms.
Step 2: (b) Median
In order: 230, 236, 241, 247, 252, 259, 268, 275. With 8 scores the median is the mean of the 4th and 5th: ms.
Step 3: (c) Range
ms.
Step 4: (d) Sample standard deviation
The squared deviations from 251 are 1, 100, 289, 441, 576, 64, 16 and 225, which sum to 1712. Then ms. On a calculator, enter the data and read (Casio) or (TI).
Psychology uses the sample standard deviation, dividing by , because the data are a sample used to estimate the spread of a population.
Try it yourself first, then check the working one step at a time.
Answer
Mean 251.00 ms; median 249.50 ms; range 45.00 ms; sample standard deviation 15.64 ms.
Method
Mean: total ÷ number of scores.
Median: order the scores; the middle one, or the mean of the middle two.
Range: highest − lowest.
Sample SD: the key, or .
Watch out for
Using the population standard deviation (), 14.63 here, instead of the sample .
Taking the median of the unordered list.
Mean reaction time for each hand. Error bars show ±1 standard deviation.
Reading the results. The dominant hand was faster on average, by 15 ms, and every one of the eight students was faster with the dominant hand. The standard deviations show that students varied a lot (about 14 to 16 ms either side of each mean), so the error bars overlap. That overlap is mostly differences between students; within each student the difference was consistent. In Topic 3 you will learn to test this with a paired t-test, which compares each student with themselves: here it gives , far below 0.05, so the difference is statistically significant and the null hypothesis is rejected for these data.
A significant result in one class of eight is still limited evidence. A larger, randomly selected sample would make the conclusion more reliable and let it generalise to other students (external validity).
Switch the error bars between SD, SE and the 95% CI. Edit a score to see what moves.
Five students' mean reaction times with the non-dominant hand were 262, 249, 271, 255 and 268 ms. Calculate the mean, to 2 decimal places.
Worked example
Evaluate a PET study of the brain and violence
Raine, Buchsbaum and LaCasse (1997) used PET to compare the brains of 41 people charged with murder who pleaded not guilty by reason of insanity with 41 control participants, each matched to one of them for age and sex. During an attention task, the prefrontal cortex of the people charged with murder used less glucose, on average, than that of the controls.
(a) Identify the type of investigation, and explain why it is not a true experiment. (b) Explain why matching for age and sex improves the study. (c) A newspaper headline reads "Brain scans show damaged brains cause murder". Evaluate this claim.
Step 1: (a) Type of investigation
It is a quasi-experiment: it compares two groups that already existed (people charged with murder and people who were not). The researchers could not manipulate the variable or randomly allocate people to groups, so it is not a true experiment.
Step 2: (b) Why matching helps
Brain activity changes with age and differs between the sexes. Matching each person charged with murder with a control of the same age and sex stops these from being confounding variables, so a difference in glucose use is not simply due to one group being older or having more men.
Step 3: (c) Evaluate the headline
The study found a difference in average prefrontal activity, not that it causes murder. Other differences between the groups could explain it (head injuries, drug use, stress of the trial, mental illness), and low activity could be a result of their life experiences rather than a cause. The sample was also a particular group (pleading insanity), so it may not apply to other people who commit violence. The headline overstates the evidence.
Evaluating a claim means weighing it against what the evidence can show: the type of investigation sets the limit.
Try it yourself first, then check the working one step at a time.
Answer
(a) A quasi-experiment: existing groups compared, no manipulation or random allocation. (b) Matching removes age and sex as confounding variables. (c) Not supported: a quasi-experiment shows a group difference, not cause; other variables could explain it; and the sample was atypical.
Method
Name the type of investigation from whether the IV was manipulated and participants randomly allocated.
Explain a control by the confounding variable it removes.
Judge a claim against what that type of investigation can show.
Watch out for
Calling it an experiment because it has two groups. Without manipulation and random allocation it cannot show cause.
Treating a group average as true of every individual in the group.
Exam tip
Name what the evidence shows, and what it does not
When you evaluate a study, name its type and say what it can conclude: an experiment can show cause; a quasi-experiment, correlational study or case study can show a difference or a relationship, but not cause. Then give one specific limitation (sample size, how participants were chosen, a confounding variable) and how to improve it.
Summary
Brain evidence comes from case studies, experiments, quasi-experiments, correlational and observational studies; only experiments show cause.
In an experiment the IV is manipulated and the DV measured; both must be operationalised. The null hypothesis predicts no effect, the alternative predicts one.
Repeated measures removes participant differences between conditions; counterbalancing controls order effects; standardised procedures remove variation in treatment.
Ethical principles (informed consent, voluntary participation, withdrawal, confidentiality, debriefing) protect participants, as early brain research often did not.
Describe data with the mean and median (centre) and the range and sample standard deviation (spread), to 2 decimal places.
Finished the lesson?
Try the 2 quick checks above and the lesson completes itself, or mark it complete now.