UCSD: Cogs 14a - Midterm 2

Barrera- Winter 202

Nominal scale of measurement

Categories that have no inherent order

• categories are just labels

• there is no inherent order to the categories

Ordinal scale of measurement

Categories that have an order

• the difference between each category is not equivalent

• simply depict the order of variables and not the difference between each of the variables. These scales are generally used to depict non-mathematical ideas such as frequency, satisfaction, happiness, a degree of pain etc. It is quite straightforward to remember the implementation of this scale as 'Ordinal' sounds similar to 'Order', which is exactly the purpose of this scale.

Interval scale of measurement

Categories(units) that are ordered and each is equivalent in size

• no true zero (true zero = absence of what is being measured)

Ratio scale of measurement

Ordered, equivalent categories (units) with a true zero

Likert Scale

A special type of interval ratings scale; carefully worded, equivalent intervals, middle value is a middle response, treated as interval data

Descriptive statistics

statistical procedures used to describe data and can be used to make inferences

• central tendency

• dispersion

Central tedency

the average value of the data

Dispersion

how close spread out the data is

Mode

The score that occurs most frequently in a distribution.

Median

The middle point (50% of scores are above and 50% are below the median)

Mean

the arithmetic average of a distribution, obtained by adding the scores and then dividing by the number of scores

Range

the number of possible scores in a dataset

• tells you how spread out the data are but not how they are distributed

Standard Deviation

the average distance a score falls from the mean

• tells you about the spread and distribution of the data

Variance

σ^2 or s^2, a measurement of the spread between numbers in a data set

68-95-99.7 rule

in a normal model, about 68% of values fall within 1 standard deviation of the mean, about 95% fall within 2 standard deviations of the mean, and about 99.7% fall within 3 standard deviations of the mean

Correlation

the relationship between variables

Regression

the relationship between an outcome variable (DV) and one or more predictor variables (IVs) for that outcome

• technically an IV is a variable that is manipulated by the researchers but it is common for people to treat non-manipulated variables (eg subject variables)

Positive Correlation

as x increases, y increases

Negative Correlation

as x increases, y decreases

Scattergrams

figures used to visualize the relationship between two variables

• each point indicates the x- and y-value of an observation

Correlation coefficient

Measures how strong the data fits a linear pattern, ranging from -1 to 1

• the polarity (+/-) indicates if it's a positive or negative relationship

• the value indicates how strong the relationship is

• the scale of measurement determine the statistical test (and what is it called)

• Pearson's r for interval and ratio data

• Spearman's p for ordinal data

Bivariate Correlations

correlations between 2 values

• correlational coefficient: Pearson's r or Spearman's p (rho)

• = range from -1 to +1

Multivariate correlations

correlations between more than 2 values

• correlational coefficient: R

= ranges from 0 to 1 (0 means no correlation, 1 means perfect correlation)

Regressions

statistical techniques for understanding how. changes in the IV (x) influence the DV (y)

Simple Linear Regression

Predicts y (the DV) based on a linear relationship to x (the IV)

Multiple linear regression (multivariate regressions)

Predicts y based on the relative contributions of each x (each IV)

• y = a + bx1 + bx2 + bxn

Nonlinear regressions

Predicts relationships based on curves rather than a single line

Error(error variance)

the amount an observation differs from its expected value. The expected value is based on the population

Fixed variables

assumed to be measured without error; the values is one study are the same as the values of another study

Random Variables

values that depend on outcomes of random phenomenon (they have error associated with them); will differ between samples and studies

Comparing Groups

(differences between groups)/(difference within groups) = (effect of iv + error variance)/(error variance)

Parameter

a characteristic of a population

Parametric tests

used if the DV is interval or ratio data and the data is normally distributed - called parametric bc there is an assumption about the distribution of the parameter (ie the DV)

Non-parametric tests

used if the DV is nominal or ordinal, or if the data is not normally distributed

• they are called non-parametric bc there is no assumption about the distribution of the parameter

• uses chi square test x^2

• common test for comparing groups

T-test

a parametric test that is used when there are only two groups being compared

Chi-square test

A significance test(non-parametric test) used to determine if a linear relationship exists between two variables measured on interval or ratio scales.

Modalities of Measurement

self-report, physiological, behavioral

Self-Report

a method in which people provide subjective information about their own thoughts, feelings, or behaviors, typically via surveys, questionnaire or interview

Pros and Cons of Self-Report

Pros: direct, fast, easy; sound reasonable(high face validity)

Cons: distortions; socially-desirable responding

Physiological Measurements

involves monitoring a respondent's involuntary responses to marketing stimuli via the use the following: single

• single-unit recordings, EEG, ERP, fMRI, EMG, EDA, HRV, eye-tracking

Pros and Cons Of Physiological Measurements

Pros: Objective measurements

Cons: equipment might be expensive, equipment might be disruptive to natural behavior, mapping to constructs

Single unit (single cell) recordings

•Place small electrode outside of the neuron. It records sudden voltage changes(when the neuron fires an action potential)

•Great spatial resolution

•Great temporal resolution

•Invasive

•Can (mostly) only be done with animals

Electroencephalography (EEG) & event-related brain potentials (ERP)

•Record electrical activity at the scalp: brainwaves (synchronized and summed postsynaptic pyramidal cell activity)

•Direct measure of brain activity

•High temporal resolution

•Poor spatial resolution

•Different neural oscillations (EEG) are associated with different activities (e.g., delta waves & deep sleep)

•Different ERP components associated with cognitive processes (e.g., N400 and semantic retrieval)

•Good for addressing questions about timing and neural processes but not locations

Functional Magnetic Resonance Imaging (fMRI)

•Indirect measure of brain activity

•Assumes changes in blood flow index

changes in neural activity

•BOLD (Blood Oxygen Level Dependent contrast)

•High spatial resolution

•Poor temporal resolution

•Good for addressing questions about where brain activity differs

Pros and Cons of Behavioral Measurements

Pros: Generally inexpensive and easy to administer; Many have a long history of use

Cons: Some behaviors can be difficult to elicit

Correlational studies

Explore the relationship between variables but no variable is manipulated by the experimenter

• doesn't provide causal information

Pilot Studies

short runs of the experiment where you check forpotential problems so that you can fix them

Measures of Dispersion

used to indicate how spread out the data is

• range

• standard deviation

• variance

normal distribution

a mean of zero and follows the 68-95-99.7 rule

Pearson's r

for interval and ratio data

Spearman's p (rho)

for ordinal data

Causation

• correlation doesn't always imply causation

• regressions can be used to address questions about predictability and causal inference

• although regressions can be used to test causality, using a regression does not entail that causal relationships are being tested

Correlation vs. Regression

From correlation we can only get an index describing the linear relationship between two variables; in regression we can predict the relationship between more than two variables and can use it to identify which variables x can predict the outcome variable y.

Comparing Variability

• when comparing groups, you want to know if the variability (error) within a group is smaller than the variability (error) between groups

• comparing the variability within and between groups in your sample is the basis for making inferences about the population

Response Time

Mental chronometry

• assumes that longer RTs are associated with additional information processing

• Many processes are involved in making a response

• perceptual

• conceptual (representational, semantic, etc)

Subject Variable

an experience or characteristic of a research participant that is not of primary interest but nonetheless may influence study results and thus must be accounted for during experimentation or data analysis. Examples include age, marital status, religious affiliation, and intelligence