Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
99 Cards in this Set
 Front
 Back
an idea constructed by the researcher to explain observed events (a rational idea that is used to explain a phenomenon); an operational definition (explanatory idea about internal states that generate behavior)

construct


How well the study’s results support the theory or
constructs behind the research; Whether the theory supported by the findings provides the best available explanation of the results 
construct validity


How do we maximize construct validity?

The researcher should use clearly stated
definitions and carefully build hypotheses on solid, well‐validated constructs which have received support from numerous other studies; have a clear definition of the constructs of interest; verify that there is a good match between constructs and operations used to represent them 

Defined as the degree of confidence that the
inferences about study outcomes based on statistical tests are correct 
statitical conclusion validity


What are the factors affecting validity types?

measurements must be reliable and valid
useful measures are sensitive useful tests are specific should avoid ceiling and floor effects researcher bias  

refers an effect whereby data
cannot take on a value higher than some "ceiling." 
celing effect


refers an effect whereby data
cannot take on a value higher than some "floor." 
floor effect


How can researcher bias be lessened?

using designs where the researcher is blind to conditions
comparisons crosschecked for consistency in recordings 

An ________ _________ is a measure of the
“influence” of a journal • Measures the ___________ with which the "average article" in a journal has been cited in a given period of time • Journal Impact Factors are produced by .....? • Impact factor available thru (1) _______________ ____________ (2) on a journal’s _________ _________ 
impact factors
frequency Thomson Institute for Scientific Information University Libraries; journal's website 

Journals which publish on more specialized
topics will have _______ impact factors than those which publish on more general topics 
lower


Other factors influencing impact factors?

Date of publication
– Size of journal – Average number of citations – Number of review articles – Type of field 

We’ve now learned about ________ of evidence
and ___________ of evidence in terms of: – Study design (experiment, quasi‐experiment, etc.) – Validity (internal, external, construct, etc.) • The highest quality, strongest and/or most innovative articles will often be the most highly influential – Journal impact factors! 
quality; strength


What section of a research paper summarizes data and statistical analysis? It uses _____________ statistics and includes tables and figures.

results section
descriptive 

_____________ statistics summarize, simplify,
and describe measurements – Measures of “_______ __________: mean, median, mode – Measures of ___________: range, variance, standard deviation 
descriptive
central tendency variability 

____________ statistics help researchers
interpret the data What are some examples? 
inferential
T‐tests, ANOVA, ANCOVA, Wilcoxon sign‐rank test, Tukey’s HSD, etc. 

The appropriateness of the descriptive and
inferential statistics for data affect _____ _________ __________ 
statistical conclusion validity


What are the different scales of measurement?

Nominal Scale
• Ordinal Scale • Interval Scale • Ratio Scale 

Naming scale
– Each number reflects a category What are some examples? We can only look at _____________, can't ____ or __________ 
nominal
men/women, type of hearing loss (conductive, sensorineural, mixed); desire to upgrade to AuD (yes/no); code as 1, 2, 3, etc.; phoneme production (correct/incorrect); diagnostic category (stutterer/nonstutterer) frequencies; add/subtract 

Scale indicating rank order
– Reflects the order, but not the amount What are some examples? We can calculate the _________ rankin but not much more. 
ordinal
ranked severity groups, socioeconomic status (low‐, middle‐, upper class), rank in class, stimulus complexity (easy, moderate, difficult) mean 

can only be represented
by integers (whole numbers like 1, 2, 3…) – Nominal and ordinal data theoretically capable of taking on fractional units of measurement – decimals (sometimes rounded) – Interval and ratio data 
discrete variables
continuous variables 

Scale with equal intervals
–The scale indicates amount, but with no zero point What are some examples? We can .... 
interval scales
temperature on the Celsius scale, most assessments (TOLD, PPVT‐R, CELF) Can add subtract, but typically not multiply or divide 

Scale that fits the number system well
– Includes equal intervals and a true zero ‐ “zero means zero” ; "score" data Examples...? We can... 
ratio scale
7 Ratio Scales • Scale that fits the number system well – Includes equal intervals and a true zero ‐ “zero means zero” • Examples: time, distance, frequency, weight, volume, etc., vowel duration, sound frequency, air pressure, number of misarticulations, diadochokinetic rate, speech intelligibility score Perform any mathematical equation 

__________: Experiments, comparative studies
____________: Correlational studies, survey research, retrospective research _______: Observational research, case studies, Interview research* ________: Interview research*, narrative research, testimonials *Strength is higher if structured interviews are used 
strong
medium low lowest 

2‐D graphs have:
– An_________ (abscissa) – A ________ (ordinate) 
xaxis
yaxis 

Frequency or grouped frequency can be
represented as either: ___________ Bar graph showing a count on the y‐axis – _______ __________: Line graph showing a count on the yaxis 
histogram
frequency polygon 

shows the %
or proportion of scores that fall within each range of values – Similar to histogram, but y‐axis is different 
relative frequency distribution


whole pie is 100%; pieces of pie
represent fractions of 100% for each category 
pie chart


Some distributions are ______________ (right side is the
mirror image of the left side) – E.g., ________ distributions: “bell curve” • Other distributions are not symmetric – ___________ _________: “Tail” tends toward positive direction – ___________ ____________: “Tail” tends toward negative direction 
symmetric
normal positively skewed negatively skewed 

We may want to know where “most of the
participants fall” on the distribution – This is called the _________ ___________ of the distribution – Measures of central tendency include __________ , ____________, ___________ 
central tendency
mean median mode 

Central tendency = _______ of a distribution
• Measures of central tendency: – _______ (average) – ___________ (middle score in rank‐ordered list) – _______ (most frequent score) 
mean
median mode 

___________ – one most frequency score
__________ – two most frequency scores 
unimodal
bimodal 

We can quantify __________(or horizontal spread)
of a distribution by using.... (2) ____________ = (standard deviation)2 (square of standard deviation) ___________ _____________ = (variance) (square root of the variance) 
variability
variance & standard deviation variance standard deviation 

A common type of unimodal distribution is the
____________ distribution 
normal


To calculate the variance, follow these steps:

– Work out the mean (average of the numbers)
– Then for each number: subtract the mean and square the result (the squared difference) – Then work out the average of those squared differences This gives you the variance! – Standard deviation is the square root of the variance 

In a histogram, counts go on the ____ axis.

yaxis


__________ statistics summarize, simplify,
and describe measurements. They include measures of ____________ __________ (mean, median, and mode; and measures of _______________: range, variance, SD. 
descriptive statistics
central tendency variability 

____________ statistics help researchers
interpret the data. These include: T‐tests, ANOVA, ANCOVA, Wilcoxon sign‐rank test, Tukey’s HSD, etc. 
inferential statistics


_________ _________ take into account (a) group mean differences relative to (b) ____________ and (c)

test statistics
variability number of participants 

_____________ ______________ statistics provides a standard way of telling whether differences in means across levels of the IV(s) outweigh variability in each level relative to number of participants
• If the inferential test statistic passes a specified __________, then there is a significant effect of that variable; otherwise there is not a significant effect. 
inferential statistics
threshold 

If there is statistical __________, then there’s no
significant effect of the IV(s) If there is statistical __________, then there is a statistically significant effect of the IV(s) 
equality
difference 

What are examples of some test statistics?

t scores (ttests), z scores (ztests), F (for ANOVA)


“Cutoff value” specifies the boundary between an extreme score from a non‐extreme score.
If a test statistic is in the “_________ __________ region”, then we conclude the IV had a statistically significant effect 
extreme score region


What are the locations of the critical region boundaries for three different levels of significance? .

α = .05,
α = .01, and α = .001 

Descriptive statistics for a sample are referred to as _________ ___________.
– Mean (M), standard dev (s), variance (s2) 
sample statistics


_________ ___________are abstract or “ideal” values of sample statistics
– Mean (μ), standard dev (σ), variance (σ2) 
population parameters


What is the first step in the logic of inferential stats tests? What are the 2 possibilities? These are stated in terms of ___________ _______________ like μ, σ.

Identify & state each hypothesis.
The levels of the IV are equal The levels of the IV are not equal population parameters 

What is step 2 in the logic of inferential stats tests?

Set the criteria for a decision:
2 substeps:  determine the statistical test  determine the cutoff value for the distribution that separates normal from extreme scores 

___________ _______________ refers to how subjects are chosen to be part of a study. It refers to drawing a sample from the population in a manner that ensures that each member of the population has an equal change of being chosen for the sample.

random sampling


________________ ____________ is a requirement of a true experiment in which an experimental group and a control group are formed by the experimenter. It is achieved when each subject is just as likely to be assigned to the control group as the experimental group.

random assignment


Research that focuses primarily on theory testing is often called _________ research. _____________ research seeks to go from data directly to a realworld application.

basic
applied 

What is the college sophomore problem?

The worry that, because college sophomores are the subjects in an extremely large number of psychological investigations, the generality of the results is in question.


What is the idea that a new theory in science must make contact with previously established empirical facts? To be considered an advance, it must not only explain new facts, but also account for old ones.

principle of connectivity


The principle of __________________________ involves scientists and those who apply scientific knowledge making a judgement about where the preponderance of evidence points. It is a useful tool for the lay consumer of scientific information and particularly useful in evaluating psychological claims. It urges us to base conclusions on data that arise from a number of slightly different experimental sources.

converging evidence


A design known as the _____________ has been used to investigate the televisedviolence/aggressive behavior issue.

field experiment


The ______________________________ refers to the failure of some people to respond with help when observing another individual in an emergency situation.

unresponsive bystander phenomenom


To conclude that there is a significant causal relationship between variable A and behavior B does not mean that variable A...

is the ONLY cause of behavior B


A factor that influences behavior may have different effects when operating in conjunction with another factor compared to when it is acting alone. This is called the concept of _____________. The magnitude of one variable may depend on the level of another.

interaction


____________________ means that it is more likely than not, but does not hold true in all cases.

probabilistic trend


The prediction of outcomes based on group characteristics is called ______________________.

aggregate/actuarial prediciton


_____________________ is the tendency for people to see links between events in the past and events in the future when the two are really independent.

gambler's fallacy


What is important when dealing with multiple causation?

interactions


For IV #1... (A,B)
What does H0 say? What is the equation that illustrates this? What does H1 say? What is the equation that illustrates this? 
“The levels of the IV are statistically equal
 There is no difference between the levels” H0: uA = uB "The levels of the IV are statistically nonequal  There is a difference between the groups” H1: uA ≠ uB 

What is step 3 in the logic of inferential stats tests?

compute test statistics from the actual data


What is step 4 in the logic of inferential stats tests?

make a statistical decision and interpret the results relative to the research question.


What are the potential benefits of studying more than one IV in a single study?

time/effort/cost
possibility of looking for interactions among different variables 

An unexpected pattern of data
in which one or more condition(s) do not follow the trends seen for the individual IV’s in the rest of the study 
interaction


A study that concludes that there are very different outcomes for men and women under different drug dosages revealed the ____________ of IV of gender and IV of drug dosage. Thus, the ________ of drug dosage depends on whether the participant is male or female.

interaction


Interactions become clear when studying two IVs simultaneously. This is done using a ____________ study, or one that has more than 1 IV.

parametric


________________ is a study where there is more than one IV and there are participants in all possible pairings of levels of the IV’s. These are the norm in studies with more than 1 IV.

factorial


First, determine how many ______ there are. This will tell you how many possible ___________________ and _______________ there are. For each, there will be a separate ________________.

IVs
main effects interactions test statistic 

A single IV will have...
Two IV’s have... 
one pair of hypotheses for each DV
3 pairs of hypotheses for each DV (one relating to possible main effect of IV 1, one for main effect of IV 2, one for interaction of IV 1 and 2 

In an experient w/ one IV ; three levels (multivalent study  experiment or quasiexperiment):
Level A, Level B, Level C – H0: ““The levels of the IV are statistically equal  There is no difference among the groups” H0: uA = uB = uC – H1: “The levels of the IV are statistically nonequal  There is a diff. among the groups” H1: uA ≠ uB and/or uB ≠ uC and/or uA ≠ uC (*posthoc test needed) 
lala


1. One IV with two levels (bivalent
study  experiment or quasiexperiment): Level A and Level B – H0: “The levels of the IV are statistically equal  There is no difference between the levels” H0: uA = uB – H1: “The levels of the IV are statistically nonequal  There is a diff. between the groups” H1: uA ≠ uB 
lala


Interaction between IV #1 and IV #2:
– H0: “IV #1 affects all levels of IV #2 in a statistically equal manner” – H1: “IV #1 affects one or more levels IV #2 in a statistically unequal manner” – No equations required 
lala


Cutoff values depend on ________________________ given by:
– Number of ____________ AND/OR – Number of ________________ 
degrees of freedom
participants levels of the IV 

This type of statistical test is based on the assumption that data fit a particular shape or distribution (e.g., ‘Gaussian distribution’). What are some examples?
_____________ tests are not based on this assumption. – Examples: Chisquared test, Wilcoxon test, MannWhitney test, KruskalWallis test 
parametric
ttest, ANOVA (f) nonparametric 

If you have one IV with two levels, you use a ...
If you have one IV with three or more levels, you use a(n)... If you have two or more IVs, you use... 
ttest (or ANOVA)
ANOVA ANOVA 

What is the "extreme" region of a distribution?
Example: With df = 20, the critical t value is +/2.09. If we get t(20) = 5.73, it’s significant! (p < .0001) 
5% or smaller portion of the curve


What are the (3) types of causation?
_______________: Where a variable is necessary and sufficient to produce an effect on a DV _________________: A causal variable might be necessary for an effect on the DV, but the effect depends on another variable. ______________: A causal variable might be neither necessary nor sufficient, but its presence increases the statistical probability of an effect. 
strongest form
weaker form weak form 

Any variable that confounds the ability of the chosen sample to represent the population parameter from which it was drawn (bias in representation due to selection, affects external & internal validity __________________________

selection bias


How is selection bias minimized?

Using random assignment to the levels of IV
Also by matching participants in terms of groups Developing welldefined inclusion criteria Develop exclusion criteria to eliminate unwanted participants that might bias a sample Sampling as close to randomly as possible 

Defined as validity of the inference about
whether the causal relationship holds over people, settings, treatment variables, measurement variables, and time ________________________ 
external validity


What are the (4) types of external validity?

population validity
treatment variation validity temporal validity ecological validity 

__________ describes an all‐inclusive data set about which researchers want to draw a conclusion.
A _________ is just a subset of the population. There will inevitably be some amount of ______________, i.e., difference between the measures collected for a sample and the population it’s believed to represent 
population
sample sampling error 

This refers to the ability to generalize from the sample in a study to a larger population. The _____________ population is the larger population to whom the results are generalized. The ____________________________ population is the one that is available to the researcher.

population
target population experimentally accessible population 

How can generalization be achieved easily?

If the researcher uses random selection


This involves drawing observations from a population so that each individual has an equal chance of being selected _________________. This is different than __________________, which sets out to create equivalent groups by balancing them based on specific characteristics.

random sampling
random assignment 

____________________ is a sampling method where each individual has an equal probability of being selected, "true random selection."

simple random sampling


___________________ involves dividing a population into subgroups called strata to assure that certain segments of the population are adequately represented in a sample. ______________ means a subdivision/group.

stratified sampling
strata 

________________________ refers to the extent to which results of the experiment can be generalized across time.

temporal validity


____________________ refers to the generalizability of results across variations in treatment. It's an issue because the administration of the treatment can vary from one administration to the next.

treatment variation validity/transferability


______________________ refers to the degree to which behaviors that are observed in a study reflect the behaviors that occur in natural settings.

ecological validity


Refers to the extent that causal inferences are justified based on observed changes in a DV in response to systematic variations in an IV
The strength of depends on to what extent extraneous variables have been removed from the study. 
internal validity


An most important possible threat to internal
validity is not having a high enough constraint study design How do we minimize this? 
study design threat
random assignment manipulation to remove extraneous variables 

The fact that an individual changes/matures over time
What is the problem with this? 
Maturation
There is no control group to compare for effects of natural maturation 

Effects of previous testing on performance
What can we do to guard against this? 
testing effects
use diff questions/assessment wait a long time before reassessing 

A change in a DV due to a tx effect that is not immediately observable but tends to become increasingly observable over a span of time
To guard against this...? 
sleeper effect
plan appropriate treatment interval and track participants after study has ended 

This is any event between the beginning of a treatment and the measurement of the DV that could produce the outcome
How can we minimize this? 
history
make sure no new events occur after beginning of tx 

Refers to the fact that some individuals 'drop out' before completing a study
Ways to minimize? 
attrition
give incentives for participating, enroll new participants in the study 