Fiche de révision : Understanding Data Types and Measurement

📋 Course Outline

1. Data Types
2. Primary Data Collection
3. Secondary Data Sources
4. Qualitative Data
5. Quantitative Data
6. Subjective Data
7. Objective Data
8. Measurement Levels
9. Nominal Scale
10. Ordinal Scale
11. Interval Scale
12. Ratio Scale

📖 1. Data Types

🔑 Key Concepts & Definitions

• Primary Data: Data that is original and collected firsthand through research methods such as surveys or experiments (see section 1). It provides direct insights specific to the study’s objectives.
• Secondary Data: Existing data gathered from second-hand sources like scientific journals, reputable websites, and databases (see section 2). It is used for analysis without direct collection efforts.
• Qualitative Data: Descriptive data that captures qualities or characteristics, often gathered through open-ended questions, interviews, or observations. It requires interpretation to identify patterns and themes (see section 3).
• Quantitative Data: Numerical data that allows for statistical and mathematical analysis, including counts, scores, times, and measurements (see section 4). It is often collected using instruments.
• Subjective Data: Personal opinions or experiences that reflect individual perceptions, often influenced by personal feelings or biases (see section 5). It is not measured against external criteria.
• Objective Data: Data measured against external, standardized criteria, aiming for impartiality and consistency (see section 6). It minimizes personal bias.

📝 Essential Points

• Data types are classified based on their origin (primary vs. secondary) and nature (qualitative vs. quantitative).
• Qualitative data provides contextual insights but requires time for pattern recognition; it often involves open-ended responses or observations.
• Quantitative data enables statistical analysis and is collected through instruments, including counts, scores, and measurements.
• Data can be continuous (measurable to any degree of precision, e.g., height, temperature) or discrete (countable, e.g., number of students).
• Levels of measurement increase in precision: nominal (labels), ordinal (ranking), interval (equal intervals, no true zero), and ratio (equal intervals with true zero). These levels influence the choice of statistical tests.
• Psychological research often prefers quantitative data due to its suitability for statistical procedures, especially with ratio and interval scales.

💡 Key Takeaway

Understanding the distinctions between data types and measurement levels is essential for selecting appropriate collection methods and statistical analyses, ensuring valid and reliable research outcomes.

📖 2. Primary Data Collection

🔑 Key Concepts & Definitions

• Original Data: Data that is collected firsthand through direct research methods such as surveys, experiments, or observations. It is unique to the researcher’s study and not previously published (source content).
• Qualitative Data Collection Methods: Techniques that gather descriptive information about qualities or characteristics, including open-ended survey questions, interviews, focus groups, and observations. These methods require interpretation of patterns and themes (source content).
• Quantitative Data Collection Methods: Techniques that produce numerical data suitable for statistical analysis, such as counts, scores, times, and measurements. Instruments like scales and measurement tools are used to gather this data (source content).
• Subjective Data: Personal opinions or experiences collected through methods like interviews or open-ended questions, reflecting individual perceptions (source content).
• Objective Data: Data measured against external criteria, such as counts or measurements, that are independent of personal feelings or opinions (source content).
• Levels of Measurement (see section 7): Hierarchical classification of data precision, including nominal, ordinal, interval, and ratio scales, which influence the type of statistical analysis that can be performed (source content).

📝 Essential Points

• Primary data is gathered directly by the researcher, offering control over data quality and relevance, unlike secondary data which is sourced from existing records (source content).
• Qualitative data provides insights into context and underlying reasons, often requiring time-consuming analysis to identify patterns and themes. It is useful for exploring complex phenomena (source content).
• Quantitative data allows for statistical procedures; it is collected using instruments that produce counts, scores, or measurements, facilitating mathematical analysis (source content).
• Psychological research frequently employs quantitative data due to its suitability for statistical testing (source content).
• Data can be classified as continuous (measurable to any degree of precision, e.g., height, temperature) or discrete (countable, e.g., number of students) (source content).
• The levels of measurement (nominal, ordinal, interval, ratio) determine the analytical methods applicable to the data, with ratio data allowing for the most powerful statistical tests (source content).

💡 Key Takeaway

Primary data collection involves gathering original, firsthand information through qualitative or quantitative methods, with the choice of technique and measurement level critically influencing the analysis and insights derived.

📖 3. Secondary Data Sources

🔑 Key Concepts & Definitions

• Secondary Data: Data that has been collected previously by other researchers or organizations, and is available for analysis without the need for new data collection (see source content).
• Reputable and Authoritative Sources: Sources such as scientific journals, official websites, and established databases that are considered credible for secondary data (see source content).
• Qualitative Secondary Data: Descriptive data from existing sources that provide insights into qualities or characteristics, often used to explore context or themes (see source content).
• Quantitative Secondary Data: Numerical data obtained from secondary sources that can be subjected to statistical analysis, including counts, scores, and measurements (see source content).
• Levels of Measurement (see section 7): The classification of data based on their properties, which influences how secondary data can be analyzed, including nominal, ordinal, interval, and ratio scales (see source content).

📝 Essential Points

• Secondary data is advantageous because it saves time and resources compared to primary data collection, but it requires careful evaluation of source credibility and relevance (see source content).
• Qualitative secondary data often includes descriptions from existing reports, interviews, or observations, which need interpretation to identify patterns and themes (see source content).
• Quantitative secondary data allows for statistical analysis, making it suitable for research requiring numerical insights, such as psychological or behavioral studies (see source content).
• The choice of data depends on research needs; for example, continuous data (like temperature) are measured precisely, while discrete data (like the number of students) are countable (see source content).
• When using secondary data, understanding the levels of measurement helps determine appropriate analytical techniques, with ratio data allowing for the most powerful statistical tests (see source content).

💡 Key Takeaway

Secondary data sources provide valuable, time-saving insights for research, but require careful evaluation of credibility and understanding of data measurement levels to ensure valid analysis.

📖 4. Qualitative Data

🔑 Key Concepts & Definitions

• Qualitative data: Descriptions of the qualities or characteristics of what is being studied, often gathered through open-ended questions, interviews, focus groups, or observations. It provides insight into context and conditions, requiring time for pattern and theme identification.

• Insight for exploring context: Qualitative data helps researchers understand the underlying reasons, motivations, and contextual factors influencing phenomena, which are not easily quantifiable.

• Likert-type scales: Although primarily used for quantitative data, Likert scales (e.g., 1-5) can be part of qualitative research to gauge attitudes or perceptions, blending numerical scoring with qualitative interpretation.

• Subjective data: Personal opinions, feelings, or experiences collected through qualitative methods, emphasizing individual perspectives rather than measurable facts.

• Pattern and theme identification: The process of examining qualitative data involves coding and categorizing information to uncover recurring themes or patterns, essential for meaningful analysis.

📝 Essential Points

• Qualitative data focuses on qualities, characteristics, and contextual insights, often collected via open-ended questions, interviews, focus groups, or observations (see section 5 for subjective data).

• It requires interpretative analysis to identify patterns and themes, which can be time-consuming but provides rich, detailed understanding.

• While qualitative data is descriptive, it can include scaled responses like Likert-type scales, which offer a semi-quantitative aspect but still primarily serve qualitative analysis.

• Psychological research often favors quantitative data for statistical analysis; however, qualitative data is crucial for exploring complex phenomena, motivations, and contextual factors.

• Unlike objective data, qualitative data is subjective, relying on personal experiences or opinions, and is often analyzed through thematic coding.

💡 Key Takeaway

Qualitative data provides rich, descriptive insights into qualities and contextual factors, requiring interpretative analysis to uncover patterns and themes that quantitative data cannot capture.

📖 5. Quantitative Data

🔑 Key Concepts & Definitions

• Quantitative data: Numerical data that can be measured and analyzed statistically, allowing for mathematical computations (source content).
• Continuous data: Measurable data that can take any value within a range, such as height, weight, or temperature, often involving degrees of precision (source content).
• Discrete data: Countable data with specific, separate values, such as the number of students or cars, determined by counting (source content).
• Levels of measurement: Hierarchical categories that define the precision and nature of data, including nominal, ordinal, interval, and ratio scales (source content).
• Nominal scale: Categorizes data without any quantitative value or order, representing labels or names (source content).
• Ratio scale: Contains all properties of interval data but includes a meaningful zero point, enabling ratio calculations, such as height or reaction time (source content).

📝 Essential Points

• Quantitative data includes both subjective and objective data, with objective data being measurable through external criteria (source content).
• Qualitative data provides descriptive insights but can be transformed into quantitative data using scales like Likert-type scales (1-5) for statistical analysis (source content).
• Measurement levels increase in precision from nominal to ratio, with ratio data allowing for the most powerful statistical tests due to the presence of a true zero (source content).
• Continuous data allows for high precision and detailed analysis, whereas discrete data is limited to specific, countable values (source content).
• Understanding the levels of measurement is crucial for selecting appropriate statistical procedures and accurately interpreting data (source content).

💡 Key Takeaway

Quantitative data is numerical and can be analyzed statistically, with the level of measurement determining the precision and type of analysis possible. Proper classification into nominal, ordinal, interval, or ratio scales is essential for valid research conclusions.

📖 6. Subjective Data

🔑 Key Concepts & Definitions

• Qualitative Data: Descriptions of qualities or characteristics of what is being studied, often gathered through open-ended questions, interviews, focus groups, or observations. It provides insight into context and helps identify patterns and themes. (Source Content)

• Subjective Data: Personal experiences, opinions, or perceptions that are influenced by individual feelings or beliefs, rather than measurable external criteria. It contrasts with objective data, which is based on observable and measurable facts. (Source Content)

• Likert-type Scales: A type of measurement scale used in qualitative data collection, typically ranging from 1 to 5, to gauge attitudes or perceptions. It allows for the quantification of subjective opinions, facilitating analysis. (Source Content)

• Personal Experience or Opinion (see Objective Data): Data derived from an individual's personal perspective, which is inherently subjective and may vary between individuals. It is often used in qualitative research to explore personal viewpoints. (Source Content)

📝 Essential Points

• Qualitative data is essential for exploring context, gaining in-depth understanding, and identifying patterns or themes, but it requires time for interpretation (see Qualitative Data).
• Subjective data is based on personal experience or opinion, making it inherently variable and influenced by individual perceptions (see Objective Data).
• Quantitative measures like Likert-type scales convert subjective opinions into numerical data, enabling statistical analysis (see Qualitative Data).
• The distinction between subjective and objective data is crucial in research design, especially in psychological studies where personal perceptions significantly influence outcomes.
• Continuous data (measurable to any degree of precision) and discrete data (countable, specific values) are often used to quantify subjective responses for analysis (see Values in Data Collection).

💡 Key Takeaway

Subjective data captures personal experiences and perceptions, providing rich insights that complement objective measurements, but it requires careful interpretation due to its inherently personal and variable nature.

📖 7. Objective Data

🔑 Key Concepts & Definitions

• Primary Data: Data that is original and collected firsthand through research methods such as surveys or experiments (source content).
• Secondary Data: Existing data gathered from second-hand sources like scientific journals, reputable websites, and databases (source content).
• Quantitative Data: Numerical data that allows for statistical analysis and mathematical computation, including counts, scores, times, and measurements (source content).
• Continuous Data: Measurable data that can be quantified to any degree of precision, such as height, weight, or temperature (source content).
• Discrete Data: Countable, separate values within a range, such as the number of students or cars (source content).
• Levels of Measurement: Hierarchical categories of data precision, including nominal, ordinal, interval, and ratio scales, used to classify and analyze data appropriately (source content).

📝 Essential Points

• Objective data is distinguished from subjective data by its basis in external, measurable criteria, reducing personal bias (source content).
• Quantitative data is particularly valuable in psychological research because it facilitates statistical procedures, enabling rigorous analysis (source content).
• Continuous data involves measurements that can be refined to any degree, whereas discrete data involves countable, separate values (source content).
• Levels of measurement increase in precision from nominal (labels) to ratio (meaningful zero), affecting the types of statistical analysis that can be performed (source content).
• Examples of nominal data include categories like sex or religion; ordinal data includes rankings like star ratings or placement; interval data involves equal intervals without a true zero, such as temperature in Celsius; ratio data includes measurements with a true zero, such as height or reaction time (source content).
• Psychological measures often utilize interval and ratio scales to allow for meaningful statistical analysis (source content).

💡 Key Takeaway

Objective data provides measurable, quantifiable information that supports rigorous statistical analysis, with the level of measurement determining the precision and type of analysis possible.

📖 8. Measurement Levels

🔑 Key Concepts & Definitions

• Nominal Scale: Represents categories or labels that cannot be ordered or quantified; mutually exclusive and discrete (e.g., sex, religion). AUTHOR (date): "Categorical data that classifies objects without any quantitative value."
• Ordinal Scale: Ranks data in a meaningful order, but intervals between ranks are not necessarily equal (e.g., placement, star ratings). AUTHOR (date): "Data that can be ordered but does not specify the magnitude of difference between items."
• Interval Scale: Numerical data with equal intervals between values but no true zero point (e.g., Celsius temperature, clock time). AUTHOR (date): "Allows for the measurement of differences but not ratios, as zero is arbitrary."
• Ratio Scale: Like the interval scale but with a meaningful zero point, enabling ratio comparisons (e.g., Kelvin temperature, reaction time). AUTHOR (date): "Provides the highest level of measurement, allowing for the calculation of ratios and meaningful zero."
• Discrete Data: Countable values within a range, with separate, distinct points (e.g., number of students). AUTHOR (date): "Data that can only take specific, separate values."
• Continuous Data: Measurable quantities that can be refined to any degree of precision (e.g., height, temperature). AUTHOR (date): "Data that can be infinitely subdivided within a range."

📝 Essential Points

• Levels of measurement are ordered from least to most precise: nominal, ordinal, interval, and ratio.
• Nominal data cannot be used for mathematical operations beyond counting or mode determination.
• Ordinal data allows ranking but does not specify the magnitude of differences; statistical analysis is limited to non-parametric tests.
• Interval data supports more advanced analysis, including mean and standard deviation, but lacks a true zero, which limits ratio calculations.
• Ratio data permits all mathematical operations, including ratios and percentages, due to the presence of a true zero point.
• Psychological research often employs interval and ratio scales to facilitate statistical procedures (see AUTHOR (date)).
• The choice of measurement level affects the types of statistical analysis that can be conducted and the interpretation of data.

💡 Key Takeaway

Levels of measurement determine the type of data collected and influence the statistical methods that can be applied; understanding these levels is essential for accurate data analysis in psychology.

📖 9. Nominal Scale

🔑 Key Concepts & Definitions

• Nominal Scale: A measurement scale that categorizes data into distinct, mutually exclusive groups without any inherent order or ranking. It assigns labels or names to different categories (see AUTHOR (date)).
• Mutually Exclusive Categories: Categories that do not overlap; each data point can only belong to one category (see AUTHOR (date)).
• Discreteness of Data: Data on a nominal scale are discrete, meaning they are separate and distinct, with no intermediate values possible between categories (see AUTHOR (date)).
• Non-Quantifiable Labels: Labels or names used in nominal scales cannot be quantified or ordered; they serve only as identifiers (see AUTHOR (date)).
• Examples of Nominal Data: Includes variables such as sex (male/female), religion (Islam/Christianity/Judaism), and first language (English/Mandarin/Arabic) (see AUTHOR (date)).

📝 Essential Points

• The nominal scale is the simplest level of measurement, used primarily for classification purposes.
• It does not support mathematical operations like addition or subtraction, only counting or frequency analysis.
• Categories are mutually exclusive, meaning a data point cannot belong to more than one category simultaneously.
• The scale is useful in psychological research for categorizing variables such as gender, ethnicity, or language, where the focus is on identification rather than ranking or measurement (see AUTHOR (date)).
• Since the scale involves labels without order, statistical analysis is limited to mode, frequency counts, and chi-square tests.
• Unlike ordinal, interval, or ratio scales, the nominal scale does not imply any quantitative relationship between categories.

💡 Key Takeaway

The nominal scale categorizes data into distinct, non-ordered groups, serving as a fundamental tool for classification in research where the primary goal is to identify and count categories rather than measure or rank them.

📖 10. Ordinal Scale

🔑 Key Concepts & Definitions

• Ordinal Scale: A measurement scale that allows data to be ordered or ranked in a meaningful sequence, but the intervals between ranks are not necessarily equal (author unknown).
• Ranking: Assigning positions or orders to data points based on a specific criterion, such as first, second, third (author unknown).
• Unknown Intervals: The precise differences between ranks are not known or are inconsistent, meaning the scale indicates order but not magnitude (author unknown).
• Examples of Ordinal Data: Ratings like star reviews, class rankings, or Likert-type scales (e.g., 1-5), where the order matters but the exact difference between points is unspecified (author unknown).
• Comparison with Other Scales: Unlike nominal scales, ordinal scales provide a sense of hierarchy, but unlike interval or ratio scales, they do not specify the magnitude of difference between ranks (author unknown).

📝 Essential Points

• The ordinal scale enables researchers to rank data points, such as preferences or performance levels, without assuming equal intervals between ranks (author unknown).
• It is useful in psychological research where subjective assessments are common, such as star ratings or Likert scales, which are often treated as ordinal data (author unknown).
• While ordinal data can be ordered, the lack of equal intervals limits the types of statistical analyses that can be performed; non-parametric tests are typically used (author unknown).
• The scale's inability to specify the magnitude of difference between ranks means it is less precise than interval or ratio scales but still valuable for understanding relative positioning (author unknown).
• Examples like placement in a race or grades (A, B, C) demonstrate how ordinal data captures order but not the extent of difference (author unknown).

💡 Key Takeaway

The ordinal scale is essential for ranking data where order matters but the exact differences between ranks are unknown, making it suitable for subjective assessments and non-parametric statistical analysis.

📖 11. Interval Scale

🔑 Key Concepts & Definitions

• Equal intervals: The defining feature of the interval scale, where the difference between any two adjacent points on the scale is consistent and meaningful, allowing for precise measurement of differences (see "Interval" in levels of measurement).
• No true zero: The interval scale lacks a true zero point that indicates the absence of the measured property, meaning ratios are not meaningful (e.g., temperature in Celsius or Fahrenheit).
• Psychological measurement: Many psychological constructs, such as intelligence quotient (IQ) or certain attitude scales, are measured on an interval scale, enabling statistical analysis of differences (see "Psychological measures" in the source).
• Mathematical operations: Addition and subtraction are valid on interval scales, but multiplication and division are not meaningful due to the absence of a true zero (see "Interval" in levels of measurement).
• Examples of interval data: Temperature in Celsius or Fahrenheit, time on a clock, and certain standardized test scores where equal differences are interpretable but ratios are not (see "Interval" overview).

📝 Essential Points

• The interval scale allows for the measurement of differences with equal intervals, which facilitates statistical procedures such as mean, standard deviation, and correlation analysis.
• Unlike nominal and ordinal scales, the interval scale provides a quantifiable and comparable measure of the distance between data points, but it cannot support ratio calculations because of the lack of a true zero (see "Interval" in levels of measurement).
• Temperature in Celsius or Fahrenheit exemplifies the interval scale because the difference between 10°C and 20°C is the same as between 30°C and 40°C, but 0°C does not mean 'no heat'.
• Psychological measurements often utilize interval scales to assess constructs like intelligence or attitudes, enabling meaningful comparison of differences (see "Psychological measures" in the source).
• The absence of a true zero means that statements like "twice as hot" are not valid on an interval scale, unlike on a ratio scale where zero indicates the absence of the property measured.

💡 Key Takeaway

The interval scale measures data with equal intervals between points but lacks a true zero, making it suitable for analyzing differences and patterns, especially in psychological and physical measurements like temperature and time.

📖 12. Ratio Scale

🔑 Key Concepts & Definitions

• True Zero Point: The defining feature of the ratio scale, indicating an absolute absence of the measured property, allowing for meaningful ratio comparisons (e.g., height, weight, reaction time). AUTHOR (date): "The ratio scale possesses a true zero point, which enables the calculation of ratios between measurements."

• Mathematical Operations: All arithmetic operations—addition, subtraction, multiplication, and division—are valid and meaningful on ratio scale data, facilitating advanced statistical analysis (e.g., calculating mean, standard deviation). AUTHOR (date): "Ratio scale data supports comprehensive mathematical computations due to its true zero."

• Comparison with Interval Scale: Unlike the interval scale, which lacks a true zero, the ratio scale's zero point signifies the complete absence of the property being measured, making ratios meaningful. AUTHOR (date): "The key distinction between ratio and interval scales is the presence of a true zero, allowing for ratio calculations."

• Examples of Ratio Data: Quantitative measures such as height, weight, reaction time, and length, where zero indicates none of the property exists. AUTHOR (date): "Measurements like height and weight exemplify ratio scale data due to their true zero points."

📝 Essential Points

• The ratio scale allows for the most precise and meaningful quantitative analysis because it includes a true zero point, enabling ratio comparisons (e.g., a person who weighs 80kg is twice as heavy as someone who weighs 40kg).

• All statistical operations are valid on ratio data, making it suitable for parametric tests and advanced analysis techniques.

• It is distinguished from the interval scale by its true zero, which signifies the complete absence of the property measured, unlike the Celsius temperature scale where zero does not mean 'no heat'.

• Examples such as height, weight, reaction time, and length are typical ratio scale data, providing the foundation for meaningful mathematical and statistical interpretation.

💡 Key Takeaway

The ratio scale's true zero point allows for meaningful ratio comparisons and comprehensive mathematical operations, making it the most informative level of measurement for quantitative data analysis.

📊 Synthesis Tables

⚠️ Common Pitfalls & Confusions

1. Confusing primary and secondary data sources; assuming secondary data is always reliable.
2. Misinterpreting qualitative data as purely subjective without analytical rigor.
3. Overlooking the importance of measurement levels when analyzing data.
4. Assuming all data types are interchangeable; neglecting the specific analysis methods.
5. Ignoring the hierarchical nature of measurement scales (nominal to ratio).
6. Using qualitative data for statistical analysis without proper coding.
7. Overgeneralizing findings from qualitative data without considering context.
8. Failing to evaluate the credibility of secondary data sources.
9. Misclassifying data as objective when it is subjective, or vice versa.
10. Neglecting the distinction between continuous and discrete data in measurement.

✅ Exam Checklist

• Know the difference between primary and secondary data, and their respective collection methods.
• Understand Creswell’s definition of qualitative research and its focus on descriptive, interpretative data.
• Be able to distinguish between qualitative and quantitative data, including their collection techniques.
• Recognize the importance of measurement levels: nominal, ordinal, interval, and ratio, and how they influence analysis.
• Know the key authors and their concepts: Patton on qualitative data, Denzin & Lincoln on qualitative methods.
• Identify suitable data types for statistical analysis versus thematic analysis.
• Understand the hierarchical nature of measurement scales and their implications.
• Be familiar with the advantages and limitations of secondary data sources.
• Recognize common pitfalls such as misclassification of data types and overreliance on secondary data.
• Know the characteristics and collection methods of qualitative data, including open-ended questions and observations.
• Be able to differentiate between subjective and objective data, with examples.
• Master the concept of measurement levels and their impact on data analysis.