Data recording, analysis and presentation — OCR A-Level Psychology
Test yourself on Data recording, analysis and presentation with OCR A-Level practice questions.
A week of Premium from your first lesson · Then the free plan continues · No card, no charge
Data recording, analysis and presentation explained
Data handling involves collection, analysis, and presentation.
Read the full explanation
Collection procedures include sampling methods, designing questionnaires, or conducting observations. Data analysis follows a logical sequence: record raw scores, identify the level of measurement, summarise using descriptive statistics, present findings in a suitable graph, and select an inferential test to interpret against a critical value. Mathematical demands include calculating percentages, fractions, ratios, mean, median, mode, and range, converting between standard and decimal form, and rounding to significant figures. Loftus and Palmer (1974) illustrates this: Experiment 1 calculated mean speed estimates across five verb conditions, while Experiment 2 tallied 'yes'/'no' broken glass responses across three conditions into a frequency table, analysed using a chi-square test.
Raw data: design of raw data recording tables; use of raw data recording tables; standard and decimal form; significant figures; make estimations from data collected.
A raw data recording table holds untreated individual results: one row per participant, one column per condition, units in headings, and a clear title. Its use is strictly for initial data capture, keeping totals and averages in a separate summary table. Standard form writes a number as a value $a$ where $1 \le a < 10$ multiplied by a power of ten (e.g., $4.5 \times 10^{-2}$), keeping small reaction times readable. Decimal form is the ordinary written version (0.045). Significant figures are counted from the first non-zero digit, so 0.04567 seconds becomes 0.046 seconds to two significant figures. Estimation involves rounding data collected to make a rough mental calculation, catching calculator errors like misplaced decimals.
Levels and types of data: nominal level data; ordinal level data; interval level data; quantitative data; qualitative data; primary data; secondary data.
The level of measurement dictates the statistical test. Nominal level data are counts in named categories, like participants obeying in Milgram (1963). Ordinal level data are ranked scores with unequal gaps, such as rating scale totals. Interval level data use fixed units of equal size, like grey matter volumes in Maguire et al. (2000); alongside normal distribution and homogeneity of variance, interval data are required for parametric tests. Quantitative data are numerical, supporting statistical analysis, while qualitative data are descriptive words preserving rich meaning. Primary data are gathered first-hand for the current aim, whereas secondary data are pre-existing, like the army recruit scores Gould (1982) reviewed.
Descriptive statistics: measures of central tendency: mode, median, mean; measures of dispersion: variance, range, standard deviation; ratio; percentages; fractions; frequency tables (tally chart); line graph; pie charts; bar charts; histograms; scatter diagram.
Descriptive statistics summarise data. Measures of central tendency include the mode (most frequent), median (middle value, resisting outliers), and mean (average of all scores). Measures of dispersion show spread: the range covers the extremes, variance measures average squared deviation from the mean, and standard deviation is its square root. Data can be expressed as a ratio, percentage, or fraction. Frequency tables (tally charts) record raw counts. Visual displays include bar charts for discrete categories, histograms for continuous data, line graphs for change over time, pie charts for proportions, and scatter diagrams for correlations, like Maguire et al. (2000) plotting hippocampal volume against time spent driving.
Inferential statistics: normal distribution curves; skewed distribution curves; probability; significance levels; using statistical tables of critical values; criteria for using a parametric test; criteria for using a specific non-parametric inferential test (Mann-Whitney U test, Wilcoxon Signed Ranks test, Chi-square, Binomial Sign test and Spearman’s Rho); understand the use of specific non-parametric inferential tests (Mann-Whitney U test, Wilcoxon Signed Ranks test, Chi-square, Binomial Sign test and Spearman’s Rho); type 1 errors; type 2 errors; symbols: =, <, <<, >>, >, ∝, ~.
An inferential test asks how likely a result is if the null hypothesis is true, accepting a probability of 0.05, or 0.01 if a false claim is costly. Test choice depends on: difference or relationship, independent or repeated design, and level of data. Chi-square takes nominal independent data, binomial sign takes nominal repeated, Mann-Whitney takes ordinal independent, Wilcoxon takes ordinal repeated, and Spearman correlates ranks. Parametric tests require interval data, normal distribution and similar variance. Decisions use critical value tables read for sample size, tails and significance level. Mathematical symbols (=, <, <<, >>, >, ∝, ~) express these relationships, while normal and skewed distribution curves illustrate data spread.
Methodological issues: representativeness; generalisability; reliability: Internal, External, Inter-rater, Test-retest, Split-half; validity: Internal, Face, Construct, Concurrent, Criterion, External, Population, Ecological; demand characteristics; social desirability; researcher/observer bias; researcher/observer effect(s); ethical considerations, including the British Psychological Society’s Code of Ethics and Conduct: Respect – informed consent, right to withdraw, confidentiality; Competence; Responsibility – protection of participant, debrief; Integrity – deception.
Representativeness and generalisability dictate if findings apply beyond the sample. Reliability is consistency: internal (split-half) and external (test-retest, inter-rater). Validity is accuracy: internal, external (population, ecological), and measurement (face, construct, concurrent, criterion). Participants may show demand characteristics or social desirability. Investigators might introduce researcher/observer bias (subjectivity) or researcher/observer effect(s) (presence alters behaviour). Ethics follow the British Psychological Society’s Code of Ethics and Conduct: Respect (informed consent, right to withdraw, confidentiality), Competence, Responsibility (protection of participant, debrief), and Integrity (deception).
Your focus
- Describe procedures for the collection of data, including sampling and measurement design.
- Calculate descriptive statistics, percentages, and convert between standard and decimal forms.
- Present data accurately using appropriate graphs with fully labelled axes and titles.
Show all 20 objectives
- Interpret the results of data analysis in the context of the original research hypothesis.
- Design a raw data recording table with appropriate rows, columns, and units.
- Convert numerical data between standard form and decimal form.
- Round collected data to a specified number of significant figures to make estimations.
- Differentiate between nominal, ordinal, and interval levels of data.
- Compare the strengths and limitations of primary and secondary data.
- Identify whether a given dataset represents quantitative or qualitative data.
- Calculate measures of central tendency and dispersion from a dataset.
- Construct appropriate graphical displays, including frequency tables, histograms, and scatter diagrams.
- Convert descriptive statistics between ratios, fractions, and percentages.
- Identify the criteria for selecting specific parametric and non-parametric inferential tests.
- Interpret statistical tables of critical values to determine significance levels and avoid Type 1 and Type 2 errors.
- Apply mathematical symbols (=, <, <<, >>, >, ∝, ~) correctly when reporting statistical results.
- Distinguish between internal and external reliability, and internal and external validity, applying specific types to research scenarios.
- Define and evaluate face, construct, concurrent, criterion, population, and ecological validity.
- Evaluate the impact of demand characteristics, social desirability, and researcher/observer bias and effects (due to presence) on research findings.
- Apply the British Psychological Society’s Code of Ethics and Conduct to assess the ethical implications of psychological studies.
Data recording, analysis and presentation exam tips
Marking Points
- Describe data collection procedures, such as selecting a sample, designing self-report measures, or setting up observational categories.
- Outline the data analysis sequence: record raw scores, identify the level of measurement, calculate descriptive statistics, draw a graph, and conduct an inferential test.
- Perform required calculations accurately, including percentages, fractions, ratios, measures of central tendency and dispersion, standard/decimal conversions, and significant figures.
- Interpret statistical findings by explaining what the calculated figure reveals about the participants or the research hypothesis, as interpretation earns significant credit.
- Anchor analysis to core studies: Loftus and Palmer (1974) calculated means for speed estimates (Exp 1) and used a chi-square test for nominal broken glass data across three conditions (Exp 2).
- When drawing a graph, ensure both axes are accurately labelled with the variable and its unit, and provide a clear, descriptive title.
- The design of raw data recording tables requires a row for each participant, a column for each condition, units in the column headings, and a descriptive title.
- The use of raw data recording tables is to capture untreated scores; summary statistics like means must be placed in a separate summary table.
- Standard form expresses numbers as $a \times 10^n$ (where $1 \le a < 10$), so $0.045$ is $4.5 \times 10^{-2}$, while decimal form writes the number out fully.
- Significant figures start from the first non-zero digit; rounding 0.04567 to two significant figures gives 0.046.
- Making estimations from data collected involves rounding values to one significant figure before calculating, providing a rough check against the final exact answer.
- Nominal level data consist of frequencies in discrete, named categories, such as counting how many participants helped or did not help.
- Ordinal level data involve ranked or rated scores where the distance between points is not guaranteed to be equal, such as a 1-10 stress rating.
- Interval level data use equal, standardised units (e.g., seconds or millimetres); they are one of the assumptions required to conduct a parametric test.
- Quantitative data are numerical and allow for statistical comparison, whereas qualitative data consist of words or narratives that provide in-depth understanding.
- Primary data are collected directly by the researcher for their specific aim, while secondary data were previously collected by someone else for a different purpose.
- Measures of central tendency include the mode for nominal data, the median for skewed ordinal data, and the mean for interval data without extreme outliers.
- Measures of dispersion include the range (highest minus lowest score) and standard deviation (average distance from the mean); variance is the standard deviation squared.
- Proportions can be calculated and expressed as a ratio (e.g., 1:4), a fraction (1/5), or a percentage (20%).
- Frequency tables (tally charts) are used to systematically record the number of occurrences of different categories or scores during data collection.
- Graphical displays must match the data: bar charts for discrete categories, histograms for continuous classes, line graphs for continuous variables, pie charts for parts of a whole, and scatter diagrams for correlations.
- State the accepted probability as 0.05, meaning a result this extreme would arise by chance about five times in a hundred, with a stricter 0.01 used where a false positive would be serious.
- Choose the test from three features of the study: difference or relationship, independent or repeated design, and level of data. Chi-square takes nominal unrelated data, the binomial sign test nominal related data, Mann-Whitney ordinal unrelated data, Wilcoxon ordinal related data, and Spearman a correlation of ranked scores.
- A parametric test requires interval data, a roughly normal distribution and similar variance across conditions, and is preferred when those hold because it is more likely to detect a real effect.
- Compare the calculated value with the critical value taken from the table for the sample size or degrees of freedom, the number of tails and the level chosen. Mann-Whitney, Wilcoxon and the sign test are significant when the calculated value is the same as or smaller than the critical value, while chi-square and Spearman are significant when the absolute calculated value is the same as or larger than the critical value.
- A type 1 error is accepting a difference that is not really there, made more likely by a lenient level, and a type 2 error is missing a real one, made more likely by a strict level; say that tightening one raises the other.
- Describe the distributions properly: a normal curve is symmetrical with mode, median and mean at the same point, a positive skew has a long tail of high scores that pulls the mean above the median, and a negative skew does the reverse.
- Define and use mathematical symbols accurately: = (equals), < (less than), << (much less than), >> (much greater than), > (greater than), ∝ (proportional to), and ~ (approximately) when reporting statistical findings.
- Representativeness is the sample sharing the characteristics of the target population and generalisability is how far findings extend beyond it.
- Name the right reliability: internal reliability (consistency within a measure, e.g., split-half), and external reliability (consistency over time or across raters, e.g., test-retest and inter-rater).
- Distinguish validity types: internal validity is whether the study measured what it claimed without confounds. Measurement validity includes face (superficial appearance), construct (underlying trait), concurrent (correlation with established test), and criterion (predictive power). External validity encompasses ecological (setting realism) and population (sample applicability) validity.
- Demand characteristics are cues that let participants work out the aim, social desirability is answering to look good. Researcher/observer bias is subjective data interpretation, whereas researcher/observer effect(s) occur when the investigator's mere presence or observation inadvertently alters participant behaviour.
- Use the four principles of the British Psychological Society’s Code of Ethics and Conduct: Respect covers informed consent, the right to withdraw and confidentiality; Competence covers working within your expertise; Responsibility covers protection of participant and debrief; and Integrity covers honesty and deception.
Examiner Tips
- 💡Always show your working for calculations; method marks can often be awarded even if a minor arithmetic slip leads to an incorrect final answer.
- 💡A calculation question is frequently followed by an explanation question on the same data, so ensure you understand what your calculated figure represents in context.
- 💡When asked for the design of raw data recording tables, ensure you include a space for every participant and do not include rows for means or totals.
- 💡An estimation question expects you to round first and calculate second, so write down the rounded figures as evidence of your method.
- 💡When justifying a level of data, refer to the properties of the measurement scale (e.g., equal units for interval) rather than just the topic of the study.
- 💡If asked to compare primary and secondary data, focus on the researcher's control over the methodology versus the time and cost efficiency.
- 💡When asked to justify a measure of central tendency, explicitly mention the level of data and whether any extreme scores (outliers) are present.
- 💡Sketching any graph (bar chart, histogram, line graph, or scatter diagram) requires fully labelled axes with units and a clear title to achieve full marks.
- 💡Test choice questions almost always ask you to justify the choice, so give the level of data and the design rather than only the name of the test.
- 💡Given a calculated and a critical value, write the comparison, the decision about the null hypothesis and the probability level in one sentence, because those are usually three separate marks.
- 💡Scenarios sometimes describe data that break the parametric criteria deliberately, so check the level and the shape of the distribution before recommending one.
- 💡Evaluation marks depend on the point being tied to the study in front of you, so quote the feature of the procedure that creates the problem.
- 💡Ethics questions often ask you to suggest an improvement, and the improvement must still allow the study to test its aim, so presumptive consent or a full debrief works better than telling participants everything in advance.
Common Mistakes
- Conflating Loftus and Palmer's experiments. Correction: Remember Experiment 1 had five verb conditions and measured mean speed, while Experiment 2 had three conditions and used chi-square for broken glass reports.
- Giving an answer to more decimal places than the raw data justify or rounding prematurely. Correction: Only round at the final step of a calculation to maintain accuracy.
- Describing a graph by merely reading numbers off it. Correction: State the overall pattern or trend and explain what it means for the hypothesis.
- Reporting a statistic without a conclusion. Correction: Always link the calculated figure back to the context of the study and the hypothesis.
- Rounding to decimal places when significant figures were asked for, so 0.04567 is written as 0.05 (two decimal places) rather than 0.046 (two significant figures).
- Designing a raw data recording table with no units in the headings, making it impossible to tell if scores are seconds or counts.
- Writing standard form with a first figure of ten or more, such as $45 \times 10^{-3}$, which is mathematically invalid as standard form.
- Assuming interval level data alone justify a parametric test, forgetting that normal distribution and similar variances are also required.
- Confusing the level of data with the type of data, offering 'qualitative' when asked if data are nominal, ordinal, or interval.
- Treating secondary data as inherently flawed, ignoring that large existing datasets can offer greater population validity than small-scale primary data.
- Drawing a histogram for discrete named categories, or leaving gaps between its bars, which incorrectly formats it as a bar chart.
- Confusing variance and standard deviation by forgetting to take the final square root when calculating standard deviation.
- Reading a scatter diagram as proof of causation, when it only displays a relationship between two co-variables.
- Getting the direction of the comparison backwards, which is the costliest error here, because a Mann-Whitney value larger than the critical value is not significant however tidy the arithmetic.
- Reading the table at the wrong number of tails, when a one-tailed value may be used only where the hypothesis states the direction of the effect.
- Writing that the results prove the hypothesis, when a significant result means the null hypothesis is rejected at a stated probability and nothing stronger.
- Mixing up the two errors, or describing a type 1 error as rejecting a true finding.
- Choosing a parametric test on ranked data because it sounds more impressive, when the criteria have not been met.
- Misinterpreting mathematical symbols, such as confusing < (less than) with > (greater than) when reporting significance levels.
- Using reliability and validity as though they were the same thing. Correction: Remember that a measure can be perfectly consistent (reliable) but consistently wrong (invalid).
- Confusing internal validity (control of extraneous variables) with measurement validities like face, construct, concurrent, or criterion validity. Correction: Ensure you treat these as distinct concepts and define them accurately.
- Confusing researcher/observer bias with researcher/observer effect(s). Correction: Bias involves interpreting data subjectively, whereas effects involve the researcher's presence influencing participant behaviour.