Conducting an investigation | AQA A-Level Psychology Revision
- Revision Notes
- Aug 5
- 30 min read
Updated: 6 days ago
For 7182 specification, first teach in September 2025
AQA A-Level Psychology | Free Revision Notes
Estimated study time: 65 minutes
This Conducting an investigation A-Level Psychology revision page explains how a research plan is put into practice. You will learn how researchers follow a standardised procedure, collect and record data accurately, protect participants throughout the study and respond to unexpected risks. You will also examine how information and communication technology can support recruitment, stimulus presentation, measurement, data storage and communication. Effective conduct is essential because even a carefully designed investigation can produce unreliable or invalid findings if its procedure is applied inconsistently.
Learning Objectives 🎯
By the end of this revision page, you should be able to:
Follow a standardised psychological research procedure.
Prepare materials, equipment and recording systems before data collection.
Give clear and consistent participant instructions.
Collect quantitative and qualitative data accurately.
Apply ethical principles throughout an investigation.
Respond appropriately to participant distress, withdrawal and unforeseen risks.
Reduce demand characteristics and investigator effects.
Use information and communication technology where appropriate.
Protect confidential research data.
Record procedural deviations and missing data clearly.
Revision Notes 📚
Conducting psychological research
Conducting an investigation means carrying out the research plan and collecting the data.
This stage begins after the researcher has decided on:
the aim and hypothesis;
the research method;
the target population and sample;
the sampling method;
the experimental design, where relevant;
the operationalised variables;
the materials and procedure;
control procedures;
ethical safeguards;
risk-management measures.
These decisions are developed in designing an investigation.
During data collection, the researcher must make sure that the agreed plan is followed consistently and ethically.
A well-conducted investigation should produce data that are:
relevant to the aim;
recorded accurately;
obtained through a consistent procedure;
collected without inappropriate influence from the researcher;
protected from unauthorised access;
suitable for later analysis.
Why the conduct of an investigation matters
A strong research plan does not guarantee strong evidence.
Problems can still arise if researchers:
give different instructions to different participants;
change the procedure during data collection;
record results inaccurately;
influence participants unintentionally;
ignore withdrawal requests;
expose participants to avoidable harm;
lose or disclose confidential data;
use faulty equipment;
enter data incorrectly;
fail to record unexpected events.
These problems could reduce:
reliability;
validity;
replicability;
the ethical acceptability of the research.
The main stages of conducting an investigation
A practical sequence is:
Prepare the setting, materials and equipment.
Confirm that the approved procedure can be followed safely.
Check participant eligibility.
provide appropriate information.
Obtain valid consent.
Assign confidential participant codes.
Give standardised instructions.
Allocate conditions or orders as planned.
Carry out the research task.
Record the data accurately.
Monitor participant wellbeing and risk.
Deal appropriately with withdrawals or unexpected events.
Debrief participants.
Store the data securely.
Document any procedural deviations.
Prepare the data for analysis without altering the original record.
Preparing to collect data
Final checks before the first participant
Before data collection begins, the researcher should check that:
ethical approval or appropriate permission has been obtained;
the procedure matches the approved research plan;
participant information and consent materials are ready;
the debrief is prepared;
equipment is working;
materials are complete;
instructions are clear;
recording sheets are correctly labelled;
researchers have been trained;
foreseeable risks have been controlled;
the testing area is suitable;
confidential data can be stored securely.
A preparation checklist reduces the likelihood that an important step will be forgotten.
Using a pilot study
A pilot study is a small-scale trial of the planned investigation.
It may reveal:
unclear instructions;
ambiguous questions;
unsuitable task length;
equipment problems;
ineffective manipulations;
difficulties recording responses;
overlapping behavioural categories;
unexpected ethical concerns;
possible risks;
floor or ceiling effects.
The researcher can then improve the procedure before collecting the main data.
Pilot work is covered in pilot studies.
Preparing the research setting
The setting should support the planned level of control.
For example, a laboratory experiment may require:
the same room for all participants;
consistent lighting;
controlled noise levels;
equipment positioned in the same way;
minimal interruptions;
a clear place for the researcher to sit;
private completion of sensitive questionnaires.
A field investigation may not permit the same degree of control, but the researcher should still record relevant situational factors.
Equipment checks
Equipment should be tested before each session where appropriate.
This could include checking:
timers;
computers;
headphones;
microphones;
cameras;
response buttons;
questionnaires;
stimulus files;
internet access where required;
battery levels;
data-storage capacity.
A researcher should also prepare a response to likely equipment failure.
For example:
If the audio recording stops during a trial, the trial will be ended, marked as incomplete and repeated only if this is permitted by the original procedure.
The researcher should not improvise an unplanned solution that treats participants inconsistently.
Checking measurement units
The researcher should confirm that measurements are recorded in consistent units.
For example:
response time in milliseconds;
task duration in seconds;
distance in centimetres;
recall score as the number of words correctly recalled;
questionnaire response using the specified rating scale.
Mixing units can make the data difficult or impossible to interpret correctly.
Following a standardised procedure
What is a standardised procedure?
A standardised procedure is a procedure in which relevant features are kept consistent across participants and conditions.
Standardisation may include:
the wording of instructions;
the order of procedural stages;
the materials used;
the time allowed;
the testing environment;
the method of recording responses;
the scoring criteria;
the debriefing process.
The purpose is to reduce unintended variation.
Why standardisation is important
Standardisation helps to improve:
reliability;
replicability;
fairness between participants;
control of extraneous variables;
accuracy of comparisons between conditions.
If one group receives more detailed instructions than another, any difference in performance could be caused by the instructions rather than the independent variable.
Standardisation does not mean every study is identical
The appropriate degree of standardisation depends on the research method.
A laboratory experiment may be highly standardised.
A structured interview can use:
identical questions;
the same order;
consistent prompts.
An unstructured interview allows more flexibility, but the researcher should still apply agreed ethical and recording procedures consistently.
A naturalistic observation may involve an unpredictable environment, but observers can still use:
the same behavioural categories;
the same sampling technique;
the same recording sheet;
the same observation duration.
Standardised instructions
Standardised instructions should explain:
what the participant must do;
how responses should be given;
how long the task lasts;
whether questions may be asked;
how to indicate withdrawal;
any relevant safety information.
The instructions should be:
clear;
concise;
appropriate for the participant group;
free from unnecessary clues about the hypothesis;
delivered in the same way to relevant participants.
Written instructions
Written instructions can improve consistency because every participant receives the same wording.
For example:
You will have two minutes to study the word list. The list will then be removed. After a one-minute distractor task, write down as many words as you can remember. You will have two minutes for recall.
This is clearer than telling each participant informally what to do.
Researcher script
A researcher may use a script covering:
greeting the participant;
checking consent;
introducing the task;
answering permitted questions;
giving the withdrawal reminder;
starting and stopping the task;
delivering the debrief.
A script reduces differences caused by one participant receiving more encouragement or explanation than another.
Timing
The researcher should apply timing consistently.
For example, every participant might receive:
two minutes to study a list;
one minute for a distractor task;
two minutes for recall.
A timer or computer program may help make this more accurate.
The researcher should not give extra time to a participant simply because they appear close to finishing.
Standardised materials
All participants in the same condition should receive equivalent materials.
These may include:
identical instructions;
the same questionnaire;
the same word list;
the same images;
the same recording;
the same response sheet;
the same equipment.
If different materials are required, they should be matched appropriately.
For example, two word lists used in a repeated measures study should be similar in:
number of words;
word length;
familiarity;
difficulty.
Standardised scoring
A scoring system should be agreed before data are collected.
For example, a memory score might be:
One mark for each word recalled exactly from the original list. Repeated words and words not on the list receive no mark.
This reduces the possibility that researchers award scores differently.
Procedural checklist
Researchers may use a checklist during each session.
Procedure stage | Completed |
Eligibility confirmed | \(\square\) |
Information provided | \(\square\) |
Consent obtained | \(\square\) |
Participant code assigned | \(\square\) |
Standard instructions given | \(\square\) |
Correct condition or order used | \(\square\) |
Data recorded | \(\square\) |
Withdrawal reminder given | \(\square\) |
Debrief completed | \(\square\) |
The checklist should support the procedure rather than distract the researcher from participants’ wellbeing.
Applying the experimental design
Independent groups
In an independent groups design, different participants complete each condition.
During the investigation, the researcher should:
allocate participants according to the planned procedure;
avoid placing participants into groups based on preference;
use the same instructions in both conditions;
keep relevant situational variables consistent;
record the condition completed by each participant accurately.
If random allocation is planned, it must be genuinely random rather than based on researcher judgement.
Repeated measures
In a repeated measures design, the same participants complete every condition.
The researcher should:
follow the planned condition order;
apply counterbalancing accurately;
record which order each participant completed;
maintain equivalent intervals between conditions;
avoid revealing unnecessary information about the comparison;
monitor fatigue and boredom.
For two conditions, counterbalancing may use:
$$A\rightarrow B$$
and:
$$B\rightarrow A$$
The number of participants completing each order should be kept as balanced as possible.
Matched pairs
In a matched pairs design, participants are paired on a relevant characteristic.
The researcher should:
apply the matching criteria consistently;
record which participants form each pair;
allocate pair members to different conditions using the planned method;
preserve the pairing in the data record.
A participant should not be rematched informally after the researcher sees their results.
Randomisation
Randomisation may be used to determine:
stimulus order;
trial order;
questionnaire order;
condition allocation;
selection from a sampling frame.
The method of randomisation should be planned and recorded.
A computer-generated random sequence can support consistency, but the researcher should preserve the sequence used.
The main control techniques are covered in control procedures.
Collecting data accurately
What does accurate data collection mean?
Accurate collection means that the recorded data reflect what participants actually said, selected or did.
Accuracy depends on:
clear operational definitions;
suitable measurement tools;
consistent units;
careful observation;
objective scoring;
complete recording;
accurate data entry.
The researcher should not change a result because it seems unexpected or inconsistent with the hypothesis.
Recording data at the time
Where possible, data should be recorded as the response occurs.
This reduces reliance on memory.
For example:
an observer records each behaviour when it occurs;
response-time software records the result automatically;
questionnaire answers are saved as participants submit them;
an interviewer records the response or makes an agreed audio recording.
Retrospective recording may result in omissions or distortions.
Participant codes
Participants should normally be assigned a code rather than having their name written on the main data sheet.
For example:
$$P001,\ P002,\ P003$$
The code allows the researcher to:
connect responses from the same participant;
preserve paired data;
process withdrawal requests;
avoid displaying names in the data set.
Where a separate list links names to codes, it should be stored securely and separately.
Recording raw data
Raw data should be preserved in its original form.
Examples include:
each participant’s test score;
the exact answer to an open question;
each observed behaviour;
the original response time;
the category selected.
The researcher should not replace raw scores with means before saving the individual results.
Summary statistics can be calculated later.
Quantitative data
Quantitative data are numerical.
Examples include:
response time;
number of words recalled;
frequency of a behaviour;
questionnaire rating;
number of correct responses.
The researcher should record:
the value;
the unit;
the participant code;
the condition or category;
any relevant trial information.
Qualitative data
Qualitative data are non-numerical and expressed in words.
Examples include:
open-question responses;
interview accounts;
descriptions of experience;
written documents.
Qualitative responses should be recorded accurately rather than reduced immediately to the researcher’s interpretation.
For example, an interviewer should not replace:
“I found the task confusing at first, but it became easier.”
with:
“Participant liked the task.”
The replacement loses important meaning.
Coding qualitative data
If qualitative data will later be converted into categories, the researcher should:
preserve the original response;
use clearly defined coding categories;
apply the coding system consistently;
record ambiguous cases;
use more than one coder where appropriate.
Coding should not be changed simply to create a clearer result.
Recording observed behaviour
Observers should use operationalised behavioural categories.
For example:
Behavioural category | Operational definition |
Offers help | Gives task information to another participant without being asked |
Requests help | Directly asks another participant for assistance |
Interrupts | Begins speaking before another participant finishes |
Praises | Makes a positive statement about another participant’s contribution |
Observers should be trained to distinguish the categories before formal data collection begins.
Event sampling
With event sampling, every occurrence of the target behaviour is recorded.
It may be appropriate for:
behaviours that are distinct;
behaviours that occur infrequently;
investigations focusing on frequency.
The observer must remain alert throughout the observation period.
Time sampling
With time sampling, behaviour is recorded at predetermined intervals.
For example, the observer might record behaviour:
every \(30\) seconds;
during the first \(10\) seconds of each minute.
A digital timer can help observers follow the intervals consistently.
Inter-observer reliability during collection
Two or more observers may record the same behaviour independently.
Their records can then be compared.
For this to be useful:
observers should watch the same events;
categories should be agreed in advance;
observers should avoid influencing one another;
their records should be made independently;
disagreements should not be hidden.
This supports the assessment of reliability covered in reliability.
Preventing data-recording errors
Clear recording sheets
A recording sheet should provide spaces for:
participant code;
condition;
score or response;
unit;
date or session;
relevant notes;
missing or withdrawn data.
A poorly structured sheet may encourage omissions or entry in the wrong column.
Consistent variable names
The same variable names should be used throughout the study.
For example, avoid using:
Score A in one file;
Memory in another;
Recall total in a third.
A consistent label such as:
Number of words recalled
reduces confusion.
Data-entry checks
When transferring handwritten or recorded data into a digital file, the researcher should check for:
transposed digits;
misplaced decimal points;
entries in the wrong row;
incorrect participant codes;
missing values;
duplicate entries;
inconsistent units.
Possible checks include:
entering the data twice and comparing the files;
asking a second researcher to verify a sample;
comparing digital totals with original recording sheets;
using data-validation rules.
Preserving original records
The original records should be retained securely where appropriate.
Corrections should not erase all evidence of the original value.
For example, if a handwritten entry is corrected, the original should remain readable with the correction clearly marked.
In a spreadsheet, a protected copy of the original imported data can be kept before cleaning begins.
Missing data
Missing data should be recorded using an agreed code or clear notation.
For example:
not completed;
technical failure;
participant withdrew;
item skipped.
A blank cell can be ambiguous because it might mean:
no response;
zero;
data not entered;
equipment failure.
The reason should be recorded where possible without revealing unnecessary personal information.
Zero is not the same as missing
A score of:
$$0$$
may be a genuine result.
For example, a participant may recall no words.
This must not be treated as missing data.
Similarly, a blank response should not automatically be changed to zero.
Recording unexpected events
Researchers should document events that might affect the results, such as:
an interruption;
equipment failure;
a participant misunderstanding instructions;
unexpected background noise;
a shortened session;
a deviation from the planned order.
The note should be factual rather than speculative.
For example:
Fire alarm began \(90\) seconds into the recall period. Session ended and no final score was recorded.
Avoiding selective recording
Researchers must not record only evidence supporting the hypothesis.
They should record:
all planned observations;
all eligible responses;
unexpected findings;
negative results;
procedural problems;
withdrawals and exclusions.
Data should not be fabricated, altered or omitted to create a preferred outcome.
Applying ethical principles during data collection
Ethics continue throughout the study
Ethics are not dealt with only by obtaining a signature before the procedure begins.
The researcher must continue to protect participants during:
recruitment;
consent;
task completion;
recording;
debriefing;
data storage;
reporting.
The main principles are covered in ethics in psychological research.
Informed consent
Before participation, individuals should normally receive enough information to decide whether they wish to take part.
The researcher should confirm that:
the participant has read or heard the information;
the information is understandable;
questions have been answered;
participation is voluntary;
consent has not been obtained through inappropriate pressure.
Consent should relate to what the participant will actually experience.
If the procedure changes substantially, the original consent may no longer be sufficient.
Checking ongoing consent
Participants may change their minds during a study.
Researchers should watch for signs that a participant:
wants to stop;
appears distressed;
is continuing only because they feel pressured;
no longer understands what is happening.
The researcher should not assume that an initial consent form removes the need to respond to these signs.
Right to withdraw
Participants should be able to leave the study without penalty.
The researcher should explain:
how to stop the procedure;
whether payment or credit is affected;
how to request removal of data;
the time limit for data withdrawal, where applicable.
The withdrawal process should be practical.
Saying that participants may withdraw is not enough if the researcher then:
challenges their decision;
asks them to justify it;
creates embarrassment;
withholds an appropriate payment;
makes data removal unnecessarily difficult.
Recording a withdrawal
Where a participant withdraws, the researcher should record:
the participant code;
the stage at which withdrawal occurred;
whether data may be retained;
whether data must be removed;
any relevant procedural information.
The researcher should not record unnecessary personal details about the reason.
Protection from harm
Researchers should monitor participants for:
physical discomfort;
anxiety;
frustration;
embarrassment;
fatigue;
distress;
signs that the task should stop.
The protection offered should be appropriate to the study.
For example, researchers using background sound should:
use a safe volume;
allow participants to stop;
check equipment;
respond immediately to discomfort.
Stop criteria
A research plan should state when participation will be stopped.
Possible stop criteria include:
the participant requests withdrawal;
the participant shows significant distress;
equipment becomes unsafe;
the procedure cannot be followed as approved;
an unexpected risk emerges;
the participant becomes unwell.
A researcher should not continue merely to avoid losing data.
Deception
Where information has been withheld or participants have been misled, the researcher should follow the approved procedure.
During data collection, the researcher should:
use only the agreed deception;
avoid inventing additional false explanations;
monitor whether the deception causes distress;
provide a full debrief afterwards;
allow questions;
explain any option to withdraw the data.
Deception should not be increased because a participant begins to suspect the true aim.
Privacy
Researchers should respect reasonable expectations of privacy.
This is especially important in:
observations;
recordings;
online research;
investigations of sensitive behaviour;
research conducted in shared spaces.
The researcher should consider whether:
consent is required;
individuals could reasonably expect not to be observed;
recorded material contains identifiable information;
other people appear in the data unintentionally.
Confidentiality
Confidentiality concerns how participant information is handled.
During the study, researchers should:
avoid discussing individual results publicly;
keep consent forms separate from data;
use participant codes;
restrict access to records;
avoid leaving screens or paperwork visible;
transfer files securely;
report data without unnecessary identifiers.
Confidentiality is not always anonymity
Anonymity means that the data cannot be linked to an identifiable participant.
If the researcher keeps a code list linking names to responses, the data are not fully anonymous, although they may be confidential and pseudonymised.
Researchers should describe the protection accurately rather than promising complete anonymity when identification remains possible.
Debriefing
After participation, the researcher should provide an appropriate debrief.
The debrief may explain:
the true aim;
the hypothesis;
any deception;
why information was withheld;
how the data will be used;
how confidentiality will be protected;
how to ask further questions;
how to withdraw data;
sources of support if the task raised concerns.
The researcher should check whether the participant has questions rather than simply handing over a sheet and ending the session.
Collecting sensitive data
When collecting sensitive information, researchers should:
ask only questions necessary for the aim;
explain why the information is needed;
allow participants to skip questions where appropriate;
provide privacy during completion;
avoid displaying responses to others;
use secure data-transfer and storage methods;
prepare a response if answers indicate significant distress or risk.
The researcher should follow the ethically approved procedure rather than improvising promises or interventions.
Managing risk during the investigation
Planned and unexpected risks
The risk assessment completed during design identifies foreseeable hazards.
During the actual investigation, the researcher must also notice unexpected problems.
Examples include:
damaged equipment;
a participant reacting badly to the task;
a change in the research setting;
an unauthorised person entering the room;
a failure in secure data storage;
a public observation becoming unsafe.
Dynamic risk management
Dynamic risk management means responding to changing circumstances while preserving participant safety and research integrity.
The response may include:
pausing the task;
stopping the session;
moving to a safer location;
replacing faulty equipment;
recording a procedural deviation;
seeking advice before continuing data collection.
Safety takes priority over completing a planned number of participants.
Incident recording
An incident record should state:
what happened;
when it occurred;
who was affected using confidential codes;
what action was taken;
whether the study continued;
whether the procedure needs review.
The record should avoid blame or unsupported interpretation.
Reducing demand characteristics
What are demand characteristics?
Demand characteristics are cues that allow participants to work out the purpose of a study or the behaviour expected of them.
Participants may then:
try to support the hypothesis;
deliberately oppose it;
monitor themselves unusually closely;
change their response to appear favourable.
Reducing demand characteristics during conduct
Researchers may:
use neutral instructions;
avoid revealing the predicted result;
avoid discussing one participant’s performance with another;
keep conditions separate where appropriate;
prevent participants from seeing completed response sheets;
use an independent groups design where comparison of conditions would reveal the aim;
provide a standardised explanation of the task.
Researchers must still satisfy ethical requirements. Reducing demand characteristics does not justify harmful or unnecessary deception.
Reducing investigator effects
What are investigator effects?
Investigator effects occur when the researcher’s behaviour or expectations influence:
participant responses;
scoring;
observation;
data entry;
interpretation.
Examples include:
encouraging one group more than another;
asking leading follow-up questions;
interpreting unclear behaviour in favour of the hypothesis;
recording borderline answers differently across conditions.
Controlling investigator effects
Possible controls include:
standardised written instructions;
researcher scripts;
objective scoring criteria;
automated stimulus presentation;
automated recording;
training observers;
using more than one observer;
keeping scorers unaware of condition where possible;
checking data entry independently.
These issues are explored in demand characteristics and investigator effects.
Conducting different types of investigation
Conducting an experiment
When carrying out an experiment, the researcher should:
check participant eligibility and consent;
follow the allocation procedure;
apply the correct condition;
use standardised instructions;
control relevant situational variables;
record the dependent variable consistently;
maintain planned timing;
monitor harm and withdrawal;
record procedural deviations;
debrief participants.
Experimental data record
A clear record might include:
Participant code | Condition | Score | Unit | Order | Notes |
P001 | Silence | \(16\) | Words recalled | AB | None |
P002 | Speech | \(11\) | Words recalled | BA | Brief corridor noise |
P003 | Silence | \(14\) | Words recalled | AB | None |
Only relevant procedural notes should be included.
Conducting an observation
When conducting an observation, researchers should:
use clearly operationalised behavioural categories;
follow the chosen time or event sampling method;
observe for the agreed duration;
record behaviour consistently;
avoid counting the same behaviour twice;
maintain privacy and confidentiality;
avoid influencing the people observed;
assess inter-observer reliability where appropriate.
Conducting a questionnaire
When administering a questionnaire, researchers should:
provide consistent instructions;
ensure participants understand how to respond;
avoid explaining questions differently to different participants;
provide privacy;
avoid looking over participants’ shoulders;
preserve skipped items as missing rather than guessing responses;
store paper and digital responses securely;
debrief participants where appropriate.
Conducting an interview
An interviewer should:
follow the planned interview structure;
use neutral language;
avoid leading participants;
ask agreed prompts consistently;
listen without showing approval or disapproval;
obtain consent for recording;
store recordings securely;
record answers accurately;
respond appropriately to sensitive disclosures.
In a structured interview, wording and order should be highly consistent.
In an unstructured interview, flexibility is greater, but ethical and recording standards remain essential.
Conducting a correlational investigation
A correlational study requires one paired set of scores for every participant.
The researcher should:
measure both co-variables accurately;
keep participant codes consistent across measures;
avoid manipulating one variable while describing the study as correlational;
preserve the pairing between scores;
record units;
avoid making causal claims during participant communication or later interpretation.
For example:
Participant code | Hours slept | Memory score |
P001 | \(6.5\) | \(14\) |
P002 | \(8.0\) | \(17\) |
P003 | \(5.0\) | \(10\) |
If the pairings are lost, the correlation cannot be calculated correctly.
Conducting a content analysis
During content analysis, researchers should:
define coding categories clearly;
select the material using the agreed procedure;
preserve the original material;
apply categories consistently;
record each coding decision;
use more than one coder where appropriate;
compare coding agreement;
document ambiguous cases.
Conducting a case study
A case study may combine:
interviews;
observations;
psychological tests;
documents;
personal records.
Researchers must:
keep a clear record of the source of each item;
distinguish fact from interpretation;
protect highly identifiable information;
obtain consent for each form of data collection;
avoid assuming that one source is automatically accurate;
record changes in procedure over time.
Using information and communication technology
What is information and communication technology?
Information and communication technology, or ICT, includes digital tools used to:
communicate with participants;
deliver research materials;
present stimuli;
record responses;
organise data;
store information;
support collaboration;
prepare data for analysis.
The AQA specification requires students to use ICT in practical research activities where appropriate.
Appropriate use of ICT
ICT should be selected because it improves the investigation, not merely because it is available.
It may support:
consistency;
precision;
efficiency;
data security;
accessibility;
accurate timing;
automated recording.
The researcher should also consider whether digital technology could reduce:
validity;
representativeness;
privacy;
reliability.
ICT in recruitment and communication
Researchers may use digital tools to:
distribute recruitment information;
arrange appointments;
provide participant information;
collect expressions of interest;
send reminders;
deliver debrief materials.
Communication should:
protect recipient details;
avoid revealing who else is participating;
use approved contact methods;
avoid pressuring people to respond;
make withdrawal procedures clear.
Online recruitment limitations
Online recruitment may create:
volunteer bias;
limited access for people without suitable technology;
uncertainty about participant identity;
difficulty confirming eligibility;
overrepresentation of particular groups.
These limitations should be considered when describing the sample.
Online questionnaires
Digital questionnaire platforms can:
present questions in a consistent order;
require or permit responses according to the design;
collect responses quickly;
reduce manual data entry;
randomise question order;
export results to a spreadsheet.
However, researchers should consider:
data protection;
where responses are stored;
whether the platform collects identifying information;
whether participants can withdraw;
whether required answers create ethical pressure;
whether the questionnaire works on different devices;
whether accidental multiple submissions are possible.
Required questions
Making every item compulsory may reduce missing data, but it may be inappropriate for sensitive questions.
Participants may need:
a “prefer not to say” option;
the ability to skip an item;
a way to leave the survey.
Ethical protection should not be sacrificed simply to obtain a complete data set.
Computerised experiments
A computer may be used to:
present stimuli for a fixed duration;
randomise trial order;
record key presses;
measure response times;
calculate task scores;
apply consistent delays between trials.
This can improve precision and standardisation.
Possible limitations
Computerised tasks may be affected by:
device speed;
keyboard differences;
screen size;
internet delay;
software errors;
participant familiarity with technology;
distractions during remote testing.
The researcher should test the program and record the equipment used where it may affect the dependent variable.
Response-time measurement
A computer can record response time more precisely than a researcher using a handheld stopwatch.
However, accuracy may still depend on:
hardware;
software;
input device;
background processes;
how the program defines the start and end points.
The researcher should not assume that digital measurement is automatically error-free.
Digital stimulus presentation
ICT may present:
images;
sounds;
videos;
written material;
timed instructions.
Digital presentation can ensure that:
every participant sees the same stimulus;
volume or duration remains consistent;
order is controlled;
trial timing is recorded.
Researchers should check:
screen brightness;
sound volume;
file compatibility;
accessibility;
whether the stimulus loads correctly.
Digital observation
Video or audio recording may allow behaviour to be:
reviewed;
coded by more than one observer;
checked for reliability;
examined in greater detail.
However, recording creates additional ethical and data-security concerns.
Researchers should obtain appropriate consent and consider:
who will access the recording;
how long it will be kept;
whether participants can be identified;
whether other people appear in the recording;
how the file will be deleted securely;
whether recording is necessary.
Spreadsheets
A spreadsheet may be used to:
enter raw scores;
label variables;
sort cases;
calculate totals;
calculate descriptive statistics;
construct tables;
identify missing values;
prepare graphs;
check data-entry ranges.
Example spreadsheet structure
Participant ID | Condition | Recall score | Order | Completed |
P001 | Silence | \(16\) | AB | Yes |
P002 | Speech | \(11\) | BA | Yes |
P003 | Silence | AB | No |
The missing score should be explained using an additional field or code rather than being silently replaced by zero.
Spreadsheet formula checking
Spreadsheet formulas can reduce repetitive calculation, but an incorrect formula may affect many results.
Researchers should check:
the selected cell range;
whether headings are excluded;
whether missing data are handled correctly;
whether percentages use the right total;
whether formulas have been copied correctly;
whether values are stored as numbers rather than text.
A small set of results can be checked manually.
Data validation
Data validation can limit entries to appropriate values.
For example, a rating scale might permit only:
$$1,\ 2,\ 3,\ 4,\ 5$$
An entry of:
$$15$$
could then be flagged.
Validation does not prove that every entry is correct, because a wrongly entered value of \(4\) would still fall within the permitted range.
Statistical software
Statistical software can support:
descriptive calculations;
data organisation;
graphs;
inferential analysis.
However, researchers must still understand:
what each variable represents;
the level of measurement;
the research design;
which test is appropriate;
how missing data are coded;
how to interpret the output.
Software can perform an inappropriate analysis accurately if the researcher selects the wrong procedure.
Secure data storage
Digital data should be stored in a way that protects confidentiality.
Possible measures include:
password protection;
access restrictions;
encrypted storage where appropriate;
secure institutional systems;
separate storage of identity information;
regular backups;
controlled file sharing;
secure deletion after the retention period.
Research files should not be left:
on an unlocked shared computer;
in a publicly accessible folder;
attached to an unsecured message;
open on a screen visible to others.
File naming
Clear file names reduce confusion.
For example:
RecallStudy_RawData_2026-08-04
is more useful than:
datafinalnew2
File names should not contain participant names unnecessarily.
Version control
Researchers should distinguish between:
raw data;
checked data;
coded data;
analysis files.
For example:
RawData_ReadOnly
DataEntry_Checked
Analysis_Copy
This reduces the risk of accidentally overwriting the original data.
Backups
Backups protect against:
device loss;
accidental deletion;
file corruption;
equipment failure.
The backup location should also meet confidentiality and security requirements.
Advantages of ICT
Appropriate ICT may:
improve timing precision;
standardise stimulus presentation;
reduce manual data entry;
make large data sets easier to manage;
support randomisation;
record responses automatically;
improve the production of tables and graphs;
make collaboration more efficient;
support secure backups;
assist reliability checks using recordings.
Limitations of ICT
ICT may introduce:
equipment failure;
software errors;
data-security risks;
digital exclusion;
differences between devices;
reduced researcher control in remote studies;
uncertainty about who completed an online task;
distractions in uncontrolled environments;
accidental duplicate responses;
loss of natural behaviour.
The researcher should evaluate whether ICT fits the method and research aim.
Monitoring reliability during the investigation
Consistency across sessions
Researchers should check that:
instructions remain unchanged;
equipment settings remain stable;
materials are used consistently;
scoring rules are followed;
observation periods remain the same length;
interview prompts are used appropriately.
A log may be kept to record each session.
Observer drift
Observer drift occurs when observers gradually change how they apply behavioural categories.
This may happen if:
definitions are forgotten;
borderline cases are interpreted differently;
observers become more lenient or strict.
Researchers can reduce observer drift through:
refresher training;
periodic agreement checks;
reference examples;
review of category definitions.
Interviewer consistency
Interviewers should avoid gradually:
changing question wording;
adding unapproved prompts;
explaining questions differently;
showing approval for certain answers.
Periodic monitoring can help preserve consistency.
Protecting validity during data collection
Following the operational definitions
Researchers should collect the exact measure stated in the design.
For example, if memory performance is operationalised as:
Number of words correctly recalled from a list of \(20\) words within two minutes
the researcher should not later include:
partially remembered words;
related words not on the list;
words recalled after the time limit.
Changing the measure would weaken operational validity and make participant scores incomparable.
Maintaining the manipulation
In an experiment, the independent variable should differ in the planned way.
For example, if the conditions are:
silence;
background speech at a standard volume,
the researcher should check that:
the speech file plays;
the volume remains consistent;
the silent room is genuinely quiet;
no additional differences are introduced.
Avoiding confounding changes
The researcher should not change several features between conditions unintentionally.
For example, if the speech condition is completed:
in a different room;
with a different researcher;
at a different time of day;
using a harder word list,
the effect of speech cannot be separated easily from these other differences.
Recording validity threats
If an extraneous event occurs, it should be recorded.
The researcher can then consider its effect during analysis and reporting.
It should not be concealed merely because it weakens the apparent quality of the investigation.
Complete worked example
Research investigation
A psychologist investigates whether background speech affects immediate word recall.
The planned study uses:
a laboratory experiment;
a repeated measures design;
an opportunity sample of \(30\) sixth-form students;
a silent condition;
a background-speech condition;
number of words recalled as the dependent variable.
Preparation
Before data collection, the researcher should:
pilot the two word lists;
check that they are similar in difficulty;
test the audio recording;
set a safe and consistent volume;
prepare consent and debrief documents;
create a counterbalancing schedule;
prepare participant codes;
test the digital timer;
prepare response sheets;
inspect the room for hazards.
Participant arrival
The researcher should:
Welcome the participant using a standard script.
Confirm eligibility.
provide the information sheet.
Answer questions consistently.
Obtain informed consent.
Explain the right to withdraw.
Assign a participant code.
Condition allocation
Participants should follow the counterbalancing schedule.
Half complete:
$$\text{Silence}\rightarrow\text{Speech}$$
The remainder complete:
$$\text{Speech}\rightarrow\text{Silence}$$
The researcher should record the order accurately.
Standardised task
In each condition:
The participant receives the same written instructions.
A list of \(20\) words is shown for two minutes.
The list is removed.
The participant completes the same one-minute distractor task.
The participant has two minutes to write recalled words.
The researcher scores one mark for each correct word.
Control
The researcher should keep constant:
room;
lighting;
seating;
instructions;
study time;
distractor time;
recall time;
scoring method;
audio volume in the speech condition.
The two word lists should be counterbalanced alongside condition order.
Accurate data collection
The data file might include:
Participant | Silence score | Speech score | Order | Notes |
P001 | \(16\) | \(12\) | Silence first | None |
P002 | \(14\) | \(11\) | Speech first | None |
P003 | \(15\) | Silence first | Audio failure |
The missing speech score for P003 should not be entered as zero.
The reason should be recorded.
Ethical conduct
During the study, the researcher should:
keep the sound at a safe volume;
monitor discomfort;
allow immediate withdrawal;
avoid displaying scores to other students;
keep consent forms separate from results;
use confidential participant codes;
provide a full debrief.
Use of ICT
ICT could be used to:
present the word lists for exactly two minutes;
play the same recording;
time each stage;
store scores in a spreadsheet;
validate that scores fall between \(0\) and \(20\);
back up the anonymised data securely.
The researcher should still check all automated timings and spreadsheet formulas.
Complete worked observation example
Investigation
Two observers record cooperative behaviour during a group problem-solving task.
Preparation
The observers should:
agree operational definitions;
practise using recorded examples;
prepare identical recording sheets;
agree whether event or time sampling will be used;
synchronise their timers;
complete an inter-observer reliability trial.
During the observation
The observers should:
remain in the agreed positions;
avoid interacting with participants;
record independently;
use the same sampling schedule;
count only behaviours meeting the operational definitions;
avoid discussing observations until recording is complete.
Ethical conduct
Participants should receive appropriate information and consent procedures.
Any recording should be:
consented to;
stored securely;
accessible only to approved researchers;
deleted according to the agreed plan.
ICT
A digital timer may signal observation intervals.
Video recording may allow later coding, but it creates additional confidentiality and privacy responsibilities.
Complete worked questionnaire example
Investigation
A researcher collects students’ views about independent revision using an online
questionnaire.
Preparation
The researcher should:
pilot the questions;
check the survey on different devices;
ensure the consent page is clear;
remove unnecessary identifying fields;
provide a “prefer not to say” option for sensitive questions;
test the data export;
prepare the debrief page.
During data collection
The researcher should:
distribute the same invitation;
avoid repeated pressure to participate;
prevent public display of email addresses;
monitor whether duplicate submissions are possible;
retain skipped items as missing data;
close the questionnaire at the planned time.
Data protection
The researcher should check:
where the platform stores responses;
who has access;
whether IP addresses or email addresses are collected;
how participants may request data withdrawal;
how data will be downloaded and stored.
A practical conduct checklist
Before collecting data
Is the procedure approved and safe?
Have the materials been piloted?
Is all equipment working?
Are standardised instructions ready?
Are participant-information documents ready?
Is the allocation or counterbalancing schedule prepared?
Are recording sheets or digital files ready?
Have researchers been trained?
Is secure data storage available?
During data collection
Has valid consent been obtained?
Is the participant code recorded correctly?
Are instructions delivered consistently?
Is the correct condition or procedure being used?
Are measurements recorded in the correct unit?
Is participant wellbeing being monitored?
Are withdrawal requests respected?
Are unexpected events recorded?
Are researcher effects being controlled?
After each participant
Is the raw data complete?
Has missing information been labelled?
Has the participant been debriefed?
Have identifiable and research data been stored separately?
Has any incident been documented?
Has equipment been reset for the next participant?
Has the correct procedural checklist been completed?
Preparing for analysis
After data collection, the researcher may:
check entries against original records;
label variables consistently;
identify missing values;
verify participant pairings;
calculate reliability where appropriate;
organise qualitative material;
prepare data tables;
calculate descriptive statistics;
select an inferential test.
The researcher should not alter the original results to make the data cleaner or more supportive of the hypothesis.
The next stages are covered in analysing and reporting an investigation.
Key Words 🔑
Key word | Student-friendly definition | How it may be used in an exam |
Conducting research | Carrying out the planned procedure and collecting evidence. | You may explain how an investigation should be implemented. |
Standardised procedure | A procedure in which relevant features are kept consistent. | You may suggest standardised instructions, timings or materials. |
Standardised instructions | The same task information given consistently to participants. | They reduce unintended variation and investigator effects. |
Researcher script | Predetermined wording used by the researcher. | It helps ensure participants receive the same explanation. |
Operationalisation | Defining a variable or behaviour so it can be measured or manipulated. | Researchers must follow the agreed operational definition during data collection. |
Raw data | The original participant responses or measurements. | Raw data should be recorded and preserved accurately. |
Participant code | A non-identifying label used to organise a participant’s data. | It helps protect confidentiality while preserving data pairings. |
Missing data | A planned measurement or response that was not obtained. | It should be coded clearly rather than automatically entered as zero. |
Procedural deviation | A departure from the planned research procedure. | Researchers should record what happened and why. |
Event sampling | Recording every occurrence of a target behaviour. | It may be used during an observation. |
Time sampling | Recording behaviour at predetermined intervals. | ICT may help observers follow the timing accurately. |
Inter-observer reliability | The extent to which observers record behaviour consistently. | Observers should record independently before comparing results. |
Demand characteristics | Cues that allow participants to guess the aim and alter their behaviour. | Neutral instructions may reduce them. |
Investigator effects | Influence of researcher behaviour or expectations on participants or records. | Scripts, training and objective scoring can reduce them. |
Informed consent | Voluntary agreement based on suitable information. | Researchers should confirm consent before participation. |
Right to withdraw | The participant’s right to stop or request removal of data. | The researcher must make withdrawal practical and free from pressure. |
Protection from harm | The responsibility to prevent avoidable physical or psychological harm. | Researchers should monitor participants and apply stop criteria. |
Confidentiality | Protecting participant information from unauthorised disclosure. | Codes, secure storage and restricted access may support it. |
Anonymity | A condition in which data cannot be linked to an identifiable participant. | It should not be promised if a code key remains. |
Debriefing | Explaining the study after participation and addressing questions or deception. | A debrief should be delivered consistently and clearly. |
Risk assessment | Identifying hazards and planning measures to reduce harm. | Researchers must continue monitoring risk during the study. |
Information and communication technology | Digital tools used to communicate, collect, organise, store or present research information. | Examples include online surveys, computers, timers and spreadsheets. |
Data validation | Rules that restrict or flag inappropriate data entries. | It may identify values outside a possible range. |
Version control | Maintaining clear, separate versions of research files. | It helps preserve raw data and track later changes. |
Observer drift | A gradual change in how observers apply behavioural categories. | Training and periodic checks may reduce it. |
Replicability | The ability of another researcher to repeat the procedure. | Detailed records and standardisation support replication. |
Hints from the Examiner Reports 💡
No lesson-specific examiner guidance was identified in the provided reports.
Common Mistakes ⚠️
Mistake: Treating the procedure as standardised merely because all participants complete the same task.
Why this is incorrect:
Instructions, timings, materials, scoring and testing conditions may still vary.
How to improve:
State exactly which features will remain consistent and how consistency will be achieved.
Mistake: Giving different explanations when participants ask questions.
Why this is incorrect:
Additional information could change participants’ understanding or reveal the hypothesis.
How to improve:
Prepare an agreed response to likely questions and use it consistently.
Mistake: Changing the procedure when a result seems unusual.
Why this is incorrect:
This creates inconsistency and may bias the evidence.
How to improve:
Follow the planned procedure and record any genuine problem or deviation.
Mistake: Assuming that computerised data collection is automatically accurate.
Why this is incorrect:
Software, hardware, formulas and device settings may contain errors.
How to improve:
Pilot the system and check automated results against known or manually calculated values.
Mistake: Recording missing data as zero.
Why this is incorrect:
Zero may be a genuine score, while missing means that no valid value was obtained.
How to improve:
Use a clear missing-data code and record the reason where appropriate.
Mistake: Recording participants’ names on every data sheet.
Why this is incorrect:
This increases the risk of unnecessary identification.
How to improve:
Use participant codes and store any identification key separately.
Mistake: Promising anonymity when the researcher keeps a participant-code key.
Why this is incorrect:
The data can still be linked to the participant.
How to improve:
Describe the data as confidential or pseudonymised unless identification is genuinely impossible.
Mistake: Assuming that consent obtained at the start cannot be withdrawn.
Why this is incorrect:
Participation remains voluntary throughout the study.
How to improve:
Monitor ongoing willingness and allow participants to stop without pressure.
Mistake: Continuing because stopping would result in lost data.
Why this is incorrect:
Participant welfare takes priority over completing the procedure.
How to improve:
Apply the agreed stop criteria and record the incomplete data appropriately.
Mistake: Giving more encouragement to one condition.
Why this is incorrect:
Researcher behaviour may affect the dependent variable and create investigator effects.
How to improve:
Use the same scripted instructions and neutral behaviour across conditions.
Mistake: Allowing observers to discuss behaviour while recording.
Why this is incorrect:
They may influence one another’s judgements, making agreement appear artificially high.
How to improve:
Observers should record independently before their records are compared.
Mistake: Changing behavioural categories after observing unexpected behaviour.
Why this is incorrect:
Changing categories during collection makes earlier and later observations difficult to compare.
How to improve:
Pilot the categories before the main study and document any necessary change before restarting or continuing under an approved plan.
Mistake: Entering only summary statistics into the data file.
Why this is incorrect:
Individual raw scores are lost and errors cannot be checked easily.
How to improve:
Preserve the raw results before calculating means, percentages or other summaries.
Mistake: Ignoring procedural interruptions.
Why this is incorrect:
An interruption could provide an alternative explanation for a participant’s result.
How to improve:
Record the event factually using the participant code and session information.
Mistake: Using online research without considering digital exclusion.
Why this is incorrect:
People without suitable technology or confidence may be less likely to participate.
How to improve:
Consider alternative access arrangements and recognise the effect on sample representativeness.
Mistake: Making sensitive online questionnaire items compulsory.
Why this is incorrect:
Participants may be forced either to disclose information or abandon the questionnaire.
How to improve:
Allow questions to be skipped or provide an appropriate non-disclosure option.
Mistake: Sharing an editable raw-data file with everyone involved.
Why this is incorrect:
This increases the risk of unauthorised access and accidental alteration.
How to improve:
Restrict access, preserve a read-only raw-data copy and share only what each researcher needs.
Mistake: Assuming ICT is appropriate for every investigation.
Why this is incorrect:
Technology may alter natural behaviour, create access barriers or introduce equipment differences.
How to improve:
Use ICT only when its benefits fit the aim, method and participant group.
Mistake: Analysing the data while collection is still continuing and then changing the procedure.
Why this is incorrect:
Later participants would experience a different investigation, and the change could be influenced by early results.
How to improve:
Follow the predetermined plan and document any approved amendment transparently.
Exam-Style Questions ✍️
Question 1
Explain what is meant by a standardised procedure in psychological research.[2 marks]
Question 2
A psychologist investigates whether background speech affects memory.
Suggest three features of the procedure that should be standardised. Explain how each feature would be kept consistent.[6 marks]
Question 3
The same participants complete two experimental conditions.
Explain how the researcher should apply counterbalancing while conducting the investigation.[3 marks]
Question 4
Two observers record aggressive behaviour during a playground observation.
Explain how they should conduct the observation so that inter-observer reliability can be assessed appropriately.[4 marks]
Question 5
A participant decides to stop halfway through a stressful laboratory task.
Explain how the researcher should respond. Refer to:
the right to withdraw;
protection from harm;
data recording.
[4 marks]
Question 6
A computer fails to record a participant’s response time during one condition.
Explain how the researcher should record and manage this problem.[3 marks]
Question 7
Explain two ways in which ICT could improve the conduct of a psychological experiment.[4 marks]
Question 8
Explain two limitations of using an online questionnaire to collect psychological data.[4 marks]
Question 9
A researcher uses a spreadsheet to store memory scores ranging from:
$$0$$
to:
$$20$$
Suggest two ways the researcher could reduce data-entry errors.[4 marks]
Question 10
A psychologist conducts an investigation into whether background music affects concentration.
Describe how the psychologist should conduct the investigation.
Your answer should include:
preparation of materials and equipment;
consent and withdrawal;
standardised instructions;
condition allocation;
accurate recording of the dependent variable;
control of investigator effects;
participant protection;
debriefing;
secure use of ICT and data storage.
[12 marks]
Comments