Home > Preview
The flashcards below were created by user
on FreezingBlue Flashcards.
When a confounding variable in the independent variable is the cause of the change in the dependent variable
attempting to find practical answers and solutions to problems
answering fundamental 'why' questions about the nature of human behavior
covariation of cause and effect
when one variable is dependent upon the other... when a cause is present, an effect occurs.
a scientific method where knowledge is based on objective observations
goals of science
- to describe behavior
- to predict behavior
scientists often use intuition for idea for research
before a study is published, it is previewed by peers, which keeps the integrity of data from being fraudulent because other members of the scientific community verify results of a study
- a systematic method for collecting, analyzing, and using information to answer questions about projects, policies and programs about their effectiveness and efficiency
- how much the program costs per participant
- how the program could be improved
- whether the program is worthwhile
- whether there are better alternatives
- if there are unintended outcomes
- whether the program goals are appropriate and useful
using scientific terms to substantiate claims
the ability to show the cause happened before the effect (differentiating between the independent and dependent variable, direction of causality)
when scientific ideas are falsified by data - showing or proving why something is false
it's been observed that 2 events are systematically parallel
why does the behavior occur
a summary of a research project, shorter than 120 words, includes info about the hypothesis, procedure, all details at the end of paper
type of idea or question that makes a statement about something that may be true on a philosophical sense
specific guess as to the outcome of the actual experiment, usually stated on the level of the operational definition... if correct, then hypothesis is true.
consists of a systematic body of ideas about a particular thought or phenomenon
predict or explain sets of hypotheses, a framework that explains a larger body of ideas
- Ethical Principles and Guidelines for the Protection of. Human Subjects of Research
- respect for persons
the degree to which one is self-serving and aware in an experiment
the benefits should outweigh the cost of an experiment
becomes particularly important when studying habits such as sexual behavior, family, or drugs
after the study subjects are told what the study was for, purpose of manipulation, and psychological evaluation
sometimes giving all the information of an experiment could interfere with the study
American Psychology Association (APA) has described the ethical code of conduct
participants in a research project should be equipped with all the information that might influence an experiment
- Institutional Review Board
- 5 people
- scientific members in said field
- at least 1 scientific member outside the field
- at least 1 nonscientific community member
the experimented debriefs a situation to a participant and then asks them how they would respond to the event
when you know that there is a third variable that affects the dependent variable instead of the change being caused a manipulated independent variable
- the adequacy of the operational definitions in identifying and representing the philosophical trait you are looking for
- must have: reliability
- must have: validity
statistical index of the strength of the relationship between variables
when the effects of a variable change over time, such as drinking enough coffee for professor is good, but drinking too much for professor is bad
- in correlational psych, the "effect"
- the changed variable when manipulating another variable
all extraneous variables are kept constant in an experiment
manipulating at least one variable and measuring a change in at least one other variable
correlational study, measures two or more variables without manipulating any
the ability of an experiment or study's results to be true in different cases
the independent variable is measured in a natural setting
- in correlational study, the "cause"
- in exp. psych, the manipulated variable
the ability to draw conclusions about causal relationships from data of an experiment
negative linear relationship
- when one variable goes up, the other goes down
- negative slope
- highly correlated
definition of the variables in order to measure changes associated to a changing variable
positive linear relationship
- increase of one variable increases the value of another variable
- highly correlated
- balances an infinite amount of confounding variables
- in an experiment can prove causation
- all subjects selected have the same chance of everything in the experiment
any event, situation, behavior that has at least 2 different conditions
does the data show a true representation of the current behavior?
the amount that the events measured are highly correlated with the actual measurement
do other studies that attempt to show the same thing highly correlated with data found?
does it prove there is a relative and significant correlation between two variables, or is it irrelevant and unconnected?
the content of the measure appears to reflect the philosophical meaning of the interested measure
internal consistency reliability
do the results show consistent results?
the extent to which two or more observers agree in their observations and recordings
the intervals between numbers are consistent but there is no 0
the amount of noise of a measurement
list of named items on a scale
list of names and the order of their ranking, but not showing the intervals between them
Pearson product-moment correlation coefficient
- a type of correlation coefficient used with interval and ratio scale
- ranging from -1 to +1
- a test of linear relationship between two variables
the extent that the results of an experiment are able to predict future behavior
a scale with equal intervals between numbers and with an absolute 0
a measurement is said to be unreliable if the actual event of being measured changes its behavior
the consistency of a measurement
when a test is split in half and both halves' results are highly correlated
when a measure is taken with good test-retest correlation, when taken at two different times and get highly correlated results, it is said to have high reliability
- newspapers, articles, statistics, documents, surveys - information that has already been collected and taken research on
- an intimate story of one experience
- highly detailed
- low external validity
- specific account of an event
a set of rules or a guide to compile, organize and categorize observations and events
researches create a coding system that raters can use to quantify the information that lies inside the documents
- researches observe in a natural setting
- paints whole picture
- hypothesis generated post hoc
- * important issue on how to categorize events: must have operational definitions to ensure objectivity
- the researches is in the environment, participating as one of the subjects
- operational definitions are measured in physical ways
- need recording equipment
- need coding system
a recording of biological data
a variable measured by coding observable behaviors as physical types of behaviors
in surveys, a method of collecting data that people report about themselves
social desirability bias
faking answers on self-report measures to seem better socially
3 common types of measures
2 ways to classify operational definitions
- categorical variables
- quantitative variables
whats another term for categorical variable?
4 types of quantitative variables?
in ordinal scales, even if one song is rated higher than another, do we know how much better it is than the other songs?
example of interval scales
- IQ scores
- Rating something from 1-10
in ratio scales, does the value of zero truly means "nothing?"
construct validity of a measure has what two aspects?
- validity of measurement
does reliability concern whether the operationalization is measuring what it is supposed to measure?
- that is construct validity
what are the 3 types of reliability assessments?
- test-retest reliability
- interrater reliability
- internal validity
can test-retest reliability be used for self-report, observational, and physiological measures?
which measure is interrater reliability most relevant to?
- OBSERVATIONAL RESEARCH:
- --naturalistic observations
- --systematic observations
- --case studies
- --archival research
evidence for reliability is a special example of an "association claim." what claim is this
- the association between one version of measure and another (internal validity)
- the association between between one coder and another (interrater reliability)
- the association between an earlier time and later time (test-retest reliability)
the two ways to measure reliability?
- correlation coefficient
can scatterplots show interrater agreement / disagreement, or test-retest reliability?
is a negative correlation for interrater reliability just as acceptable as a positive correlation?
no. negative correlation means the two observers didn't agree.
why do psychologists asks so many similar questions in internal reliability?
to ensure an average of all responses is taken as one score, canceling all confounding variables
with a philosophical idea, the question is whether the operational definition that the researches measure are representative and relevant of said idea
even though head measurement tests are reliable and have consistent reports, are they valid to describe intelligence?
no. they do not measure intelligence, and are measuring a variable that is indifferent to the variable in question.
what are the 2 subjective judgments of validity?
- content validity
- face validity
______ and ______ validity evaluate whether the measure being considered is related to a concrete outcome that it should be related to through correlational evidence.
what 2 validities check if the data is relevant, significant and purposeful between two variables?
- convergent validity
- discriminant validity
can a measure be more valid than it is reliable?
no. it must be reliable first to be considered valid data.
is it the strength of correlation or the direction of correlation that determines it there is a convergent (true) or divergent (false) validity?
strength of correlation
what is scientific progression?
theory --> hypothesis --> predictions --> experiment --> data --> conclusion
showing a relationship between two variables does not show _____ ____ _____.
cause and effect
guides research methods in psychology
curved relationship in correlation
linear correlation relationship
- not changing
- always same
what does a non-monotonic relationship suggest?
as one variable increases or decreases, the other changes in a non-linear fashion
- a variable that correlates with both the independent and dependent variables
- there is an alternative explanation for a causal relationship between the independent and dependent variables
the experimenter manipulates at least one variable and then measures at least one outcome
- an artificial environment where the independent variable is manipulated
- often uses deception
- often uses confederates
cell phones and sperm counts
- POSSIBLE EXPLANATIONS:
- stressful job --> higher cell phone use --> stress lowers sperm
- electromagnetic radiation in cell phone --> electromagnetic radiation kills sperm
- third lurking variable, alternative explanation, no correlation between cell phones and sperm depletion
touch therapy for premature infants
- hypothesis: human touch makes infants gain weight
- POSSIBLE EXPLANATIONS:
- nurses touch healthier looking babies anyway
- touch --> stimulates growth hormones --> weight gain
- after a randomized experiment proved benefit, tactile/kinesthetic stimulation may have important clinical and developmental benefits
cancer radical mastectomy treatment
- theory: cancer spreads spatially, so remove cancerous and surrounding tissue
- one large randomized experiment proved there was no benefit in radical mastectomy
blue eyes / brown eyes classroom
- Hypothesis: being included in a particular group can affect our perception of people from other groups
- Deception: kids did not know it was an experiment
- displayed in-group aggression
- theory: people display obedience when authority or encouragement is involved
- deception: learned is confederate, encourager is a confederate
- results: 65% of subjects obeyed confederates
Zimbardo's Stanford prison
- hypothesis: the brutal behavior of prison guards has more to do with social roles than personalities
- deception: role-playing
- results: had to be stopped, guards too abusive, inmates too perturbed
relationship between work-related stress and stomach ulcers
- theory: stress causes stomach ulcers
- procedure: one monkey must push lever for 8 hours to avoid shock, one monkey passive
- results: all working monkeys developed stomach ulcers and died, all passive monkeys ulcer-free
- flaws: monkeys were not randomly assigned
Marc Hauser primate body in mirror self-recognition
- results: data fraud for primates recognizing their own reflections in mirrors
- difficulties in replicating data
- students reported suspicious data fraud activity
Clark and Hatfield Study
- theory: the sex with more parenting responsibility is the limited reproductive factor for the other sex
- hypothesis: more males will agree to casual sex than females
- results: 75% of males say would sleep with woman they just met, 0% of females say would sleep with man they just met
Buss et. al.
- theory: early hormone exposure causes development of different preference mechanisms in men & women, cross-cultural factors in attraction
- hypothesis: females seek mates with abilities to provide resources, males seek beauty
- prediction: females will report higher ambition and earning capacity as more important than men report these traits, men will report beauty as more important than women report these traits
- procedure: over 37 cultures, all matched hypothesis, hypothesis correct
- possible problems with internal validity: direction of causality
Robert Cialdini's studies of persuasion
using low-ball technique, more people agree to any kind of offer
limited disclosure, make them commit, then disclose full details
costly signaling theory
- signals are regarded as more reliable and honest if they are
- --hard to fake
diddling in baboons
- theory: costly signaling theory - females rely on expensive traits in order to pick mates
- coding system: time spent together, time spent grooming
- reliability: unreported
- reactivity: animals were from zoo, so naturalistic observations were regular and common, not affecting results
- results: positive correlation of the amount of time spent together and grooming of an intense diddling greeting
Literary Digest poll - Roosevelt election
- 10 million ballots, 2 million returned
- participant pool: from telephone books and vehicle registries
- problems: selection bias, non-response bias - non-respondents may have different choice than respondents
alcohol experiment, and expectancy effects as a confounding variable
- any drinker who expected vodka showed aggression
- any drinker who expected tonic showed no aggression
- it didn't matter what drinkers actually drank. the *expectancy* of what they were drinking was the only variable to cause change in the dependent variable (aggression)
real versus sham acupuncture
results: acupuncture really works, because the results of receiving real acupuncture were higher than the results from receiving sham acupuncture
coffee shop on a hill, effects on cholesterol experiment: list the confounds, any effects of control groups, solutions.
- confounds: a shop on the hill makes an imbalanced drinkers' level of exercise, drinking more liquid per day may just lower cholesterol, no control group
- control groups: one placebo decaf, one real decaf, one real coffee, one placebo coffee
- solutions: make every participant walk up that hill once a day, make every participant drink a liquid daily
Bandura study of imitative aggression
- procedure: kid either observes adult beating toy, being passive to toy, or no observing before playing with a toy by themselves
- experimental design: matched pairs design
- hypothesis: kids who saw aggression will be more aggressive
- procedure: make kids pissed off then let them play with a toy
- flaw: making kids intentionally pissed off before experiment
- results: aggressive adult made aggressive kid. in passive adult and control group there was no difference though.
- experimental design: one group pretest-posttest design
- no problems!!!: effective example of a one group pretest-posttest design
interpreting nonequivalent control group designs
- due to no random assignment, there may be differences between experimental and control groups beyond the hypothesized cause
- key to good nonequivalent control group design is matching the control group as much as possible to the experimental group
Cohen et. al.
- theory: south as a "culture of honor"
- hypothesis: an interaction effect of regional origin and insult, with southerners reacting with more aggression than northerners
- study 1: 2 x 2 factorial design
- procedure 1: bumps and calls subjects "asshole"
- results 1: no significant effects
- study 2: same as 1 with additional dependent
- procedure 2: cortisol and testosterone
- results: southerners more affected and used violence
interrupted time series design
- multiple pretest-posttests before and after treatment, like TSST experiment saliva samples
- has control group
- the more data points, the more accurate your average will become
interrupted time series example
- historical versus current CO2 levels
- post-industrialization is unprecedented
single case design
- want to test causal effects of a manipulation on a single subject
- experimenter manipulates an independent variable
- uses interrupted time series designs
- ABA designs - base, treatment, base measures
limitations of reversal designs
- some treatment effects are difficult to reverse
- reversal designs most useful for seeing short-term treatment effects
- multiple baseline designs:
- -- introduce treatment at dif. times for dif. subjects
- -- introduce treatment at dif. times for dif. behaviors
- -- introduce treatment at dif. times for dif. situations
example of multiple baseline design
little girl traumatized by war is given treatment photos at different times of 4 therapy sessions lasting 13 days each.
statistics review - the term for assessing the probability that results could be due to chance
a smaller p-value indicates what?
a very high probability of likelihood that change is due to the manipulation of a variable
population means are equal; any differences between sample means are due to chance (random error); H sub 0
population means are not equal; H sub 1
test statistic with a probability of obtaining sample means that differ by observed amount, if population means were equal
if t is large, is the difference between the groups much bigger or much smaller than the normal variability within groups?
much bigger; therefore, two groups are significantly different from each other
if t is small, is the difference between the groups much bigger or much smaller than the normal variability within groups?
much smaller; therefore, two groups are not significantly different from each other
statistics: if two graphs' upside-down parabolas do not overlap, it is unlikely to get means this far apart by chance alone. t is ___[small/large]___ and the p-value is ___[low/high]____.
- if two graphs' upside-down parabolas do not overlap, it is unlikely to get means this far apart by chance alone.
- t is large
- p-value is low
statistics: if two graphs' upside-down parabolas overlap, it is likely to get means this far apart by chance alone. t is ___[small/large]___ and the p-value is ___[low/high]____.
- if two graphs' upside-down parabolas overlap, it is likely to get means this far apart by chance.
- t is small
- p-value is high
difference between groups (means)
variability within groups/sample size
computes p-values for main effects and interactions in factorial designs
ANOVA - ANalysis Of VAriance
extension of the t-test; computes ratio of systematic variance to error variance; similar to t-test in that it compares differences between groups to variability within groups; assesses the probability that main and interaction effects could be due to chance alone
statistics - gullibility error. solution: lower the p-value that is considered significant.
"Type 1" error
statistics - blind to a relationship. solution: run more tests to determine causal relationship.
"Type 2" error
statistics - the p-value at which we decide to reject the null hypothesis; this number is often used as a compromised between the odds of getting Type 1 and Type 2 errors
alpha = 0.5
p-values are the _______ of effect size and correlation coefficient; the lower the p-value, the greater likelihood a manipulation is the cause of a change in a dependent variable
Hierarchy of Internal Validity
- 1) True Experiments
- -- Posttest only
- -- Pretest-Posttest
- -- Repeated Measures
- -- Matched Pairs
- 2) Quasi-Experiments
- -- One group pretest-posttest
- -- Nonequivalent control group
- -- Nonequivalent control group pretest-posttest
- 3) Correlational data
true experiments have 1 important factor: what is it and why?
randomization, because it cancels all possible confounding variables, isolating only the independent variable
Say want to test whether ppt or
chalkboard lectures produce higher scores. Instructor uses ppt
in Psych 7 in one quarter but chalkboard in next quarter and compares exam
scores. What type of design? Say chalkboard scores were higher. Alternative explanations?
- nonequivalent control group design
- alt. expl.: teacher more equipped with chalk, students smarter in 2nd quarter
Say you surveyed GRE prep instructors who use chalkboard vs. ppt and then compared the average GRE scores of their classes. What type of design? Say instructors who used ppt had classes that scored higher on average. Good evidence? Alternative explanations? Was the quasi-experimental evidence better?
- correlational data
- Say instructors who used ppt had classes that scored higher on average. Good evidence? no; confounds: different teachers, different classes and material, different student intelligence
- alt. expl.: historical effects
- Was the quasi-experimental evidence better? yes; same teacher, same course material, fairly comparable students
Say drew a simple random sample of Psych 7 class and randomly assigned half to
quarter-long tutoring sessions with one TA using chalkboard descriptions and
half to quarter-long tutoring sessions with a different TA using ppt descriptions. What
type of design? Say the chalkboard group did better: good evidence? Confounds?
- posttest only true experiment
- Say the chalkboard group did better: good evidence? no; counfound: different TAs
example of a nonequivalent control group pretest-posttest design that tests whether scores on Psych 7 exams are greater with chalkboard or ppt explanations
- fall quarter give pretest to psych 7 class & record exam scores & give posttest. spring quarter give pretest to psych 7 class & record exam cores & give posttest.
- problems: teacher may be better at constructing one or the other, intelligence levels of class non-random
Say subjects came into a lab and first
took one version of a test on interpreting factorial designs. Then randomly
assigned to watch a video of: (1) instructor explaining factorial designs on a
chalkboard, or (2) same instructor explaining factorial designs with ppt slides (examples were the same in the two cases). Subjects then take second version of test on interpreting factorial designs. What type of design? Internal validity of this design? Possible problems or limitations? Say chalkboard was better: does this mean we should end ppt in classroom?
- pretest-posttest true experiment
- Internal validity of this design? unrealistic representation of lectures in a video module
- Possible problems or limitations? this one professor may be better at teaching in one style or the other; this one subject may be better taught in one style or the other
- Does this mean we should end ppt in classroom? no; video is confound of reality; external validity of professor's comfort to teach in one style or the other, and external validity to other subjects, are both poor
content validity is to what extent a measure represents all facets of a given social construct or philosophy or theory or idea. give an example.
- a depression scale may lack content validity if it only assesses the effective dimension of depression but fails to take into account the behavioral dimension
- a depression scale may lack content validity if it only assesses the effective dimension of depression, but fails to take into account the behavioral dimension.
face validity is to what extent people agree a recording superficially appears to measure. face validity assesses whether the test "looks valid" to the examinees who take it, the administrative personnel who decide on its use, and other technically untrained observers. Does the test "look like" a measure of the construct of interest? give an example.
if you prepare a test to measure whether students can perform multiplication, and the people you show it to all agree that it looks like a good test of multiplication ability, you have shown the face validity of your test.
2 limitations to face validity to participants or subjects in an experiment
- social desirability bias
- demand characteristics
As part of the informed consent process, an investigator tells participants she is interested in whether a physical exertion might harm recall memory. Half the participants are randomly assigned to stand on one foot while the other half are randomly assigned to sit, while both groups read from a list to try remembering as many words as possible. What acts as a confounding variable that may impair inferences about the relationship between memory and physical exertion?
The participants' awareness of which conditions they are being assigned to
A researcher wants to know if balance is better in the morning or afternoon. The same participants are randomly assigned to balance a book on their head in the morning, and then in the afternoon. What acts as a confound?
The order effects of time of day
In the case study of a stroke patient with damage to Broca's area, what was clearly demonstrated?
Language mechanisms must be somewhat modular since some skills can be impaired while others are normal
In an experiment of L-dopa or saline placebo injected into patients in multiple trials, what experimental design was used?
It is likely that the 1936 Literary Digest poll incorrectly forecast the results of the presidential electron because of what?
a sampling frame that was unrepresentative of the population
_______ indicates the amount of deviation in results obtained from a sample compared to the true population value.
"Do you agree that male teenage drivers are more dangerous than female teenage drivers?" is an example of a(n) ______ question.
A researcher begins testing one group of children when they are 3 years old and retests them when they are 6 and 9 years old. Over the same time period, the researcher administers the same test to another group of children when they are 9 years old, then retests them when they are 12 and 15. This researcher has used the _____ method.
Which offers a possible solution to the problem of experimenter bias?
a) staged manipulations
b) automated recording of dependent measures
c) a pretest-posttest design
d) none of the above
automated recording of dependent measures
A researcher installed a hidden camera at a traffic light in order to test the hypothesis that men would run a red light more than women. This study is an example of _____.
a) naturalistic observation
b) systematic observation
c) a field experiment
d) archival research
e) a and c
The study of conformity provides an example of a _____ manipulation
A researcher wants to know whether wearing sunglasses improves driving performance. In a complete counterbalanced repeated measures design, presence or absence of sunglasses would be accomplished by
a) randomly assigning half the drivers to a sunglasses condition and half to a no-sunglasses condition
b) having all drivers first drive without sunglasses and then with sunglasses
c) having half the drivers first drive without sunglasses and then with sunglasses and then having the other half first drive with sunglasses and then without sunglasses
d) none of the above
having half the drivers first drive without sunglasses and then with sunglasses and then having the other half first drive with sunglasses and then without sunglasses
The "bogus pipeline" is a technique designed to
avoid a social desirability bias
Which of the following is an advantage of using repeated measures design over posttest only designs?
a) a repeated measures design requires more participants than posttest only designs
b) a repeated measures design is less likely to detect the effect of the independent variable
c) a repeated measures design reduces variability due to individual differences among the participants
d) a repeated measures design decreases the likelihood of an order effect
e) c and d
- a repeated measures design reduces variability due to individual differences among the participants
- a repeated measures design decreases the likelihood of an order effect
An investigator would like to know whether a subject's resting heart rate just before exercising affects how much their heart rate increases after running on a treadmill. Which is the best research design to address this question?
a pretest-posttest experiment
An investigator hypothesized that brightness of lighting has effects on mood. To test this hypothesis, the investigator arranged two testing sessions. In one session, the investigator himself administered a mood scale to 50 participants in a bright room. In a second testing session, the same investigator administered the same scale to 50 other participants in a dim room. Participants were randomly assigned to one of the two conditions. What may threaten the internal validity of the study by acting as a confounding variable?
a) the brightness of lighting
b) the investigator's knowledge of his own hypothesis
c) participants' mood just before they arrived at the study
d) All of the above
e) a and b
the investigator's knowledge of his own hypothesis
In the brightness/mood study, what experimental design has been used?
posttest only experiment
A confounding variable is devastating to an experimental design because it ______ for the results.
makes possible alternative explanations for the results
Respondents expressing greater life satisfaction on sunny days vs. cloudy days are an example of:
an assimilation effect.
Researchers engaged in naturalistic observation generate hypotheses
Probability sampling techniques are most important for
a) generalization of results from a sample to a population
b) recruiting participants for experiments
c) obtaining statistically significant results
d) internal validity
??? generalization of results from a sample to a population ???
An advantage of participant over nonparticipant observation is that
the observer may be able to experience events in the same way as natural participants
Which is an advantage of pretest-posttest designs over posttest-only designs? With a pretest-posttest design, one can:
a) test whether the experimental groups had equal scores on the dependent variables before the experimental manipulation takes place
b) more easily disguise the hypothesis being tested
c) use more time to conduct the experiment
d) minimize demand characteristics
e) b and d
test whether the experimental groups had equal scores on the dependent variables before the experimental manipulation takes place
Students are asked to evaluate the food in the school cafeteria on 7-point scales with bipolar adjectives such as poor-good, and inexpensive-expensive. These measures represent what type of scale?
a) graphic rating
b) comparative rating
c) semantic differential
d) behavioral measure
While in a shopping mall, a marketing researcher stops and asks you which you prefer, soda in a bottle or soda in a can. Your response would be an example of a _____ measure.
Cross-sectional studies had shown that IQ tends to drop with age, starting at around age 40. Subsequent longitudinal studies have provided evidence that the cross-sectional results were due to
a) developmental changes within individuals
b) biased sampling
c) cohort effect
d) experimenter bias
- ??? cohort effect ???
- ??? developmental changes within individuals ???
What are uncovered by comparing longitudinal studies to cross-sectional studies (these are called sequential designs)?
cohort effects and developmental effects
An investigator interested in the effectiveness of various memory strategies instructs randomly assigned participants to use either imagery or repetitive silent speech to remember a list of 20 words. Two weeks later, she measures participants for how many words they recall. The imagery group remembers an average of 2.1 words, and the repetition group recalls an average of 1.9 words. She concludes there is no difference in the effectiveness of these memory strategies because the results are not statistically significant. Is this a reasonable conclusion?
No, because the two-week delay in recall may have produced a floor effect.