COGS 14A Midterm 2

non-experimental methods

are used when you can’t (or shouldn’t) manipulate something in your unit of analysis.

e.g. concerned with ecological validity or the phenomenon can’t be duplicated in a lab

examples of non-experimental methods

• field studies

• surveys and interviews

• longitudinal studies

• case studies; ‘natural experiments’

field studies

allow observations of people or animals in their natural state

surveys and interviews

can be used to gather qualitative and quantitative data

longitudinal studies

track changes in an individual or groups of people over time

case studies or ‘natural experiments’

are used to examine and characterize abilities or deficits of individuals with rare conditions or disease states; this is idiographic research

non-experimental research methods: advantages

• high ecological validity

• lack of control allows for completely unexpected observations

non-experimental research methods: disadvantages

• lack of control can allow many confounds

• open to observer bias

• with no manipulation of an independent variable, you don’t have a basis for making causal inferences

• may have ethical concerns

• may not replicate

experimental research methods: main feature

they’re typically used in situations where the independent variable has been directly manipulated by the experimenter

reaction times (RTs)

often attempt to quantify the onset of duration of some cognitive operation. sometimes called mental chronometry.

• the most widely used method in cognitive psychology

• can test sensory systems (psychophysics), motor, memory, cognitive load, attention, etc.

reaction times (RTs): theoretical assumption

RT = sensory processing time + object recognition + decision + motor response

reaction times (RTs): advantages

• cheap + non-invasive

• accepted in a variety of disciplines

• has good objectivity, reliability, and sensitivity *if used correctly

reaction times (RTs): disadvantages

• very simple. the measure may not be limited in internal validity as it may not measure the process you intend.

• doesn’t specifically indicate brain responses by itself

eye tracking *not likely on test

infrared cameras can detect eye fixations + microsaccades (tiny movements). provides measure of where the eye is focused + for how long

electroencephalography (EEG)

records electrical activity at the scalp occurring as a result of ionic currents within similarly oriented neurons in the brain.

electroencephalography (EEG): theoretical assumptions

post-synaptic potentials (EPSPs and IPSPs) are appropriate measures of internal cognitive states

event-related potentials (ERPs)

take the raw EEG data and average it over many trials. this filters out “noise” and allows for better analysis of specific components

EEG and ERP: advantages

• fantastic temporal resolution (milliseconds)

• rich data sets; lots of information

  • non-invasive + safe

EEG and ERP: disadvantages

• very poor spatial resolution of activity inside the head

  • inverse problem: infinite number of ways to model activity recorded at the scalp

• must be tightly controlled (no movement)

• hard to do outside the lab, limiting ecological validity

magnetoencephalography (MEG)

detects magnetic fields produced by the synchronized activity of neurons in the brain

magnetoencephalography (MEG): advantages *not likely on test

• excellent temporal resolution + better spatial resolution than EEG

• non-invasive + safe

magnetoencephalography (MEG): disadvantages *not likely on test

• very limited in tasks (mainly sensory + decision studies)

• all studies done in a shielded chamber; poor ecological validity

computed tomography (CT) *not likely on test

mostly used as a diagnostic tool (descriptive); uses many different X-ray images to reconstruct a 3D image of a living brain
- not used much for research; mildly invasive
- mainly highlight structural damage in brain

positron emission tomography (PET) *not likely on test

• provides a 3D image of a functional brain process

functional Magnetic Resonance Imaging (fMRI)

• tracks the flow of blood within the brain with the blood oxygenation level dependent (BOLD) signal.

• any neural activity triggers increased blood flow to replace glucose + oxygen used

functional Magnetic Resonance Imaging (fMRI): how it works

• large magnet aligns protons within brain

• brief radio pulses change orientation of protons

• protons within different tissue types return to their original orientation at different rates

• deoxygenated hemoglobin (used up) is more magnetic than oxygenated hemoglobin (going to active region)

functional Magnetic Resonance Imaging (fMRI): theoretical assumption

blood flow is a good proxy of cognitive activity

functional Magnetic Resonance Imaging (fMRI): advantages *not likely on test

• excellent spatial resolution

• considered safe, non-invasive

functional Magnetic Resonance Imaging (fMRI): disadvantages *not likely on test

• poor temporal resolution

• noisy + confined; poor ecological validity

• very expensive

magnetic resonance imaging (MRI)

• can be used to track the diffusion of water in brain tissue; can provide images of white matter; called Diffusion Tensor Imaging

cognitive neuroscience methods chart A

cognitive neuroscience methods chart B

transcranial magnetic stimulation (TMS) *not likely on test

• can be used to induce an electric current in the neurons of the cortex

• can stimulate activity in a particular region or cause temporary inactivity

• used to determine if a brain area is causally necessary for a particular behavioral function

• considered non-invasive; however, risk of seizures may be present depending on protocol

• considered an independent variable

National Commission for the Protection of Human Subjects of Biomedical

and Behavioral Research (United States)

created by the National Research Act of 1974

Belmont Report

• generated by the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research

• provided ethical principles and guidelines for research involving humans

principles outlined by the Belmont Report

1. respect for persons

2. beneficence

3. justice

respect for persons

informed, voluntary consent + protection for those needing protection

• free consent

• informed consent

• individual rights: privacy + confidentiality, maintained. violations made clear to participants

• participants who can’t decide for themselves must be protected by a responsible party

• a participant can stop the experiment at any time

• deception to be avoided; never used to cause participant to act in ways they otherwise would’ve avoided

  • if used, must be justified by nature of experiment and explained clearly during debriefing

beneficence

humans should not be harmed; aim is to maximize benefit and limit any potential harm

• risks to be kept at a minimum; if any risk is present, the benefit to society must greatly outweigh the risk

• minimal risk

• appropriate care and steps must be taken to mitigate risks as much as possible

• can be subjective + open to interpretation

  • who must benefit + how much?

justice

participant selection should be fair + benefits should accrue to everyone

• participants should be recruited fairly + equally to any study

• risks + benefits should apply equally to all participants or potential participants

  • can’t experiment on one group for the benefit of a different group

• mainly applies to the selection of participants

free consent

All participation in human studies must be voluntary and free from coercion

informed consent

Adequate information must be provided so that the participant can provide _____

minimal risk

the risk or discomfort must be comparable to what would be experienced in ordinary life

Institutional Review Boards

independent committee; its general goal is to protect human participants from physical or psychological harm

• all facilities receiving federal funding must have an IRB or equivalent

• all IRBs must have at least 5 members

  • mixed gender

  • at least 1 scientist

  • one non-scientist

  • one community member

• review all proposed research protocols at the institution and approve, approve with changes, or reject each submission

• at UCSD, the IRB operates under the Human Research Protections Program (HRPP), which ensures it follows federal, state, University policies

the Belmont report provided the ethical principles that ___ should adhere to

federally funded research

the Common Rule

the federal regulations which are used by the United States Department of Health and Human Service to provide oversight to Institutional Review Boards are codified under Title 45 of the Code of Federal Regulations, Part 46 (recently revised)

• the regulations encapsulate previous ethical guidelines, including those defined by the Declaration of Helsinki

Ethical Principles of Psychologists and Code of Conduct

issued by the APA to help identify ethical concerns in research

Institutional Animal Care and Use Committee (IACUC)

oversees all proposed research protocols for animal research on campus

• animal welfare is an important concern in research: the goal is to minimize animal pain and distress, or propose alternative methods when possible [replacement, reduction, and refinement]

• ensures compliance with federally mandated regulations through reviews of procedures and on-site inspections

• because the Belmont principles (consent, beneficence, justice) can’t be applied, this remains controversial

scientific misconduct

a violation of the regulatory or scholarly codes of ethical behavior in scientific research. may involve:

• falsifying data

• reporting/manipulating data to inaccurately reflect the results of an experiment

• excluding or suppressing data that do not support the research hypothesis

• plagiarizing another person’s work without giving appropriate credit

• including (or excluding) authors on a paper in a way inconsistent with their contribution to the work

Andrew Wakefield

a British physician who published a paper in 1998 in the Lancet purporting to find a link between the measles, mumps, and rubella (MMR) vaccine and onset of autism and bowel disease.

• a medical tribunal found that he falsified data

• had an undeclared conflict of interest: funded by a lawyer suing MMR manufacturers

• 1998 paper was retracted + lost his medical license

• measles outbreaks have continued to occur in the US

searches: if new to a topic…

see what the autocomplete gives you; this might help redirect you to specific keywords

searches: when doing a search for a paper…

use the campus internet or a VPN to use institutional subscriptions for journals

to limit Google Scholar search results…

use a boolean search, author, or time range with keywords

boolean search

• “AND” narrows (filters) the search space

• “OR” expands the search space

• NOT excludes and narrows the search space

citations

are critical in reporting results as they indicate ties to other research and also guard against charges of plagiarism.

• done in-line with the text and in the references

statistical test

we perform ___s on our measurements to arrive at a decision regarding our statistical null hypothesis

• the ____ is often determined by the scale of the data (nominal, ordinal, interval, ratio)

parameter

a value that reflects something in the entire population of interest

statistic

a value that reflects something from a sample. it can be said to estimate the population parameter.

sampling error

a difference between the population parameter and the sample statistic due to random chance.

• this is not an error as in mistake, it is due to random variability.

• unless dealing with extremely small and very uniform population, the sample statistic will (almost) always be different from the population parameter

• due to random selection of individuals from the population and inherent variability that exists in the population

• sample statistics from different samples will also be different from each other for the same reason

between-groups design

usually, the two samples have been treated with different levels of an independent variable to see how this will affect a dependent variable (when comparing two different samples from a population)

nominal measures

labels or categories that classify qualitative data. these are often names that express a difference in kind.

ordinal measures

a set of qualitative categories that are organized in an ordered (ranked) sequence

• differences of magnitude or degree of the category differentiate values. however, the differences between all values may not be directly comparable

interval measures *not likely on test

a set of ordered categories where the categories form intervals of equal size

• zero value is arbitrary

• quantitative

ratio measures

set of ordered categories where the categories form intervals of equal size

• there is a true zero

• quantitative

statistical analyses

used to (1) compare different conditions or samples and (2) generalize from sample data back to the larger population of interest

• we use the central tendency to tell us what most of the data in a sample are like

• we use the dispersion (variability) to tell us how the data in the sample vary

• can be used to compare different samples/groups

descriptive statistics

before comparing groups, compute sample statistics (estimates of population parameters) that describe our data. ways of doing this:

• measures of central tendency

• measures of dispersion

• measures of relationships

mean

the sum of all score values divided by the number of scores.

• commonly known as the arithmetic average

• uses all scores to calculate the central tendency + is the most sensitive measure of central tendency

• used in most (parametric) statistical analyses

• appropriate for interval and ratio scale data.

• not valid for nominal or ordinal scale data

• affected by outliers; extreme values “pull” the mean in one direction

median

the middle value when observations are ordered from least to greatest

• divides the distribution such that 50% of observations are at or below that value

• appropriate for quantitative data that is ordinal, interval, or ratio scale data

• not valid for nominal data

• not affected by outliers