Variables and Measurement

Variables and Measurement

Agenda

  • How numbers are collected scientifically

  • Variables

  • What is a variable?

  • Independent vs. Dependent variables

  • Functions of Data

  • Forms of Data

  • Measurement

  • How to measure

  • Measurement validity & reliability

Variables

What is a Variable?

  • Definition of a Variable: In scientific research, a variable indicates a thing (e.g., phenomena, beliefs, behaviors, attitudes, etc.) that researchers try to measure in some way.

  • Characteristics of Variables: If a variable does not vary, it is not considered a variable.

  • Constant: The opposite of a variable, a constant does not change or vary from case to case.

    • Examples of Constants:

    • Weather conditions (in a controlled study)

    • Attitude toward wearing face masks

    • Age (in a limited demographic)

    • Income (if fixed)

    • Fixed interest rate on a loan

  • Remark: There are not many constants in the real world since most traits involve variability, which is more relevant in actual experiments.

What is Data?

  • Definition of Data: Data refers to numbers that are meant to measure something.

  • Conceptual Ideas: Everything we are interested in begins as a concept, which is intangible and not inherently observable.

  • Transforming Concepts into Data: Concepts must be made empirical (observable), thus converted into data.

    • Example: Gender must be categorized into measurable data points.

    • Options include: Male, Female, Other

  • Empirical Representation: Data represents an empirical version of some aspect of the concept.

    • Examples of Concepts and Corresponding Data:

    • Concepts: “Male”, “Female”

    • Data Points:

      • Chromosomes, genitals, perceived masculinity or femininity, hairstyle choices, facial hair, clothing choices.

Types of Variables: Independent Variable

  • Independent Variable (IV): Also known as the predictor variable.

    • Function: Influences a dependent variable (DV).

    • Researcher Interest: Researchers examine how IV impacts DV.

    • Manipulation: IV is manipulated or changed by researchers in controlled experiments (not surveys).

Types of Variables: Dependent Variable

  • Dependent Variable (DV): Also referred to as the outcome variable.

    • Dependence: DV is influenced by IV(s) and “depends” on them.

    • Manipulation: DV is not directly manipulated or changed by researchers; they investigate changes in DV as a response to alterations in IV(s).

Example of Variables in Research

  • Research Focus: Kylan's dissertation explores the effect of text messages on romantic relationships.

    • Independent Variable (IV): The use of the “face blowing a kiss” emoji (😘) between a couple.

    • Dependent Variable (DV): Relationship satisfaction.

Two Functions of Data

  1. Describe (Descriptive Statistics)

    • Purpose: Gather numerical information from a sample.

    • Example: Determining the demographic of university attendees.

  2. Infer (Inferential Statistics)

    • Purpose: Utilize numerical information to make meaningful assumptions about a population.

    • Example: Assessing types of commercials and outreach materials needed to attract target population for university attendance.

    • Projection: Establishing trends beyond data, e.g., predicting the implications of current advertising strategies.

Four Forms of Data: Levels of Measurement

  1. Nominal

  2. Ordinal

  3. Interval

  4. Ratio

Overview of Forms of Data

  • Nominal Level:

    • Characterization: Categorizes and labels variables with no order.

  • Ordinal Level:

    • Characterization: Ranks categories in order, but does not offer equal intervals.

  • Interval Level:

    • Characterization: Has known, equal intervals between measures without a true zero.

  • Ratio Level:

    • Characterization: Contains a true or meaningful zero point.

Forms of Data: Detailed Explanations

  • Nominal Data:

    • Definition: Categorical or qualitative data with distinct categories.

    • Examples: Sex, gender, political party, religion, make of car, type of animal, sports team.

  • Ordinal Data:

    • Definition: Nominal properties with a ranking system, but intervals between ranks are not equal.

    • Examples: Rank in class, place in race, stages of relationships.

  • Interval Data:

    • Definition: Nominal + Ordinal properties, with ranking that includes equivalent differences between levels.

    • Examples: Strongly agree/disagree scales, quality rankings, temperature (to an extent).

  • Ratio Data:

    • Definition: Nominal, ordinal, and interval properties, with a true zero point.

    • Examples: Weight, income, percentages, temperature (to an extent).

Importance of Understanding Variables

  • Scientific Research Foundation: Variables are the most fundamental component in scientific research.

  • Definition Reiteration: Anything that researchers aim to measure, such as phenomena, beliefs, behaviors, attitudes, etc.

  • Data Dependency: Without a variable, there is no data generated.

  • Statistical Tools Relevance: Software tools in statistics apply differently depending on the type of variable; understanding definitions enhances critical evaluation of scientific research.

  • Data Relevance: Example situations such as temperature studies illustrate the practical importance of proper variable identification and classification.

Measurement

  • Definition of Measurement: The process of systematic observation and assignment of numbers to phenomena according to established rules.

Three Steps of Measurement

  1. Conceptualize IV and DV: Define what Independent Variable and Dependent Variable will be.

  2. Determine Examples of the Concepts: Compile a list of indicators for the concepts.

  3. Assign Numbers/Values to Indicators: Quantitatively assess the indicators identified.

Example of Measurement Application

  • Hypothesis Example: “Will playing violent video games increase aggression in young children?”

    • Independent Variable (IV): Playing violent video games.

    • Dependent Variable (DV): Aggression.

Step 1: Conceptualization

  • Conceptualization Definition: The development and clarification of concepts.

  • Reliable Sources: Academic research studies, government reports, and trustworthy dictionaries.

  • Example Concepts:

    • Violent Video Games: Defined by legislation as games allowing players to engage in harm towards avatars of humans.

    • Aggression: Defined as hostile or destructive behavior or actions.

Step 2: Determining Indicators of Concepts

  • Thinking About Examples: Consider games such as Call of Duty, Resident Evil, Mortal Kombat, and Fortnite.

  • Defining Aggressive Behavior Indicators: Behaviors like hitting, breaking windows, engaging in verbal assaults.

  • Multiple Indicators for Concepts: Concepts often have more than one measurable indicator; researchers select the most relevant ones for analysis.

Step 3: Assigning Numbers to Indicators

  • Process of Quantification: Indicators that can be quantified become viable variables.

  • Quantifying IV: Playing violent video games can be measured by various methods:

    • Unobtrusive Methods: Secretly observing gameplay duration using games like Fortnite, or obtaining hours logged through guardians.

    • Obtrusive Methods: Directly asking participants about their gameplay duration through surveys or interviews.

    • Example Question: “On average, how many minutes do you play Fortnite a day?”

  • Quantifying DV (Aggression): Similar processes can apply to aggression indicators.

    • Unobtrusive Measurement: Monitoring children's actions and counting aggressive incidents over time.

    • Obtrusive Measurement: Self-reported surveys asking children about their aggressive behaviors in predefined timeframes.

Evaluating Measurements

  • Key Question: What constitutes a robust measure of violent video games and aggression?

  • Criteria:

    • Validity: Does the measurement accurately reflect what it purports to measure?

    • Reliability: Does the measurement yield consistent results over repeated applications?

Validity Types

Face Validity

  • Definition: Subjective assessment of how a measurement instrument correlates with actual measurement goals.

  • Characteristics: Represents low vs. high levels of face validity without statistical tests.

Criterion Validity

  • Definition: Assesses how well a new measure correlates with a previously established validated measure.

  • Characteristics: Involves statistical tests (correlation tests) to establish validity.

    • Comparison: New, infrequently used aggression metrics vs. well-established aggression metrics.

Assessing Reliability

  • Definition of Reliability: Refers to the degree to which a measure yields approximately the same results over time.

  • Two Main Approaches:

    1. External Consistency: Test-retest method assessing stability over time.

    2. Internal Consistency: Split-half method ensuring that similar question sets evaluate the same construct sufficiently.

    • Ensures each question contributes equally to measuring the intended variable.