Transcript Steps in the Scientific Method Scientific method

Handout 1-8
1. What conclusion does the article imply?
– Homosexuals and heterosexuals differ on a biological level and that
homosexuality is caused by genetics. Point to the statements such as “suggests
a biological phenomena” and “we’ve always believed that being gay or lesbian is
not a matter of choice,” “it might explain why homosexuality is present in most
human populations.”
2. Is the conclusion warranted? No, it is a correlation.
– Alternative interpretations = 1) brain cell size  sexuality; 2) sexuality  brain
cell size; 3) distressing event  sexuality and brain cell size. Do not confuse
correlation with causation.
3. Is the title an accurate summary of the study described?
– No because it implies all gay men (even though it only tested some). All
suggests that all cells are different in gay men, but the study refers only to brain
cell nuclei. The title also suggests that the gay men are the different ones, but
the article reports gay men’s brain cell nuclei to be similar in size to those of
women. Thus, the different ones are actually the heterosexual men.
4. Can this study prove being gay or lesbian is not a matter of choice?
– The study cannot prove that it is not a matter of choice. Other variables may
contribute. Biased sample. Discuss the nature of scientific experimentation and
the inappropriateness of the term prove in science.
Steps in the Scientific Method
Scientific method - Set of
3. Form Hypothesis (assertion or prediction for
rules and procedures on how to
a behavior stated as a testable
study, observe, or conduct
proposition, usually in the form of an ifthen statement)
experiments

1.
Formulate a research
question

2.
How does self-esteem
affect academic
achievement?
Develop Theory (set of
tentative explanations of
behavior and mental
processes)
 Children with high selfesteem tend to have
high academic
achievement
 If a child has high self-esteem, than
he/she will be academically more
gifted than a child with low selfesteem
 can be verified or falsified
 examine relationships between
variables – specific factors that
are manipulated and measured
in research.
 operational definitions – a
statement of specific procedures - are used to define variables
 High self-esteem can be
operationally defined as
what a self esteem test
measures. Academic
achievement can be
operationally defined as
one’s GPA score
Population + Sample
4. Figure out population and
sample
 Population- Larger collection
of people about which we want
to generalize
 Sample - collection of subjects
used in a study
 Random Sampling every member of the
population being studied
should have an equal
chance of being selected for
the study
- Population – DHS
psychology students
- Sample – 50 randomly
sample students from
the Psych classes
Population
Sample
Samples Critical Thinking
• At the end of the first two weeks of the baseball season,
newspapers start publishing the top ten batting averages. The
leader after the first two weeks normally has a batting average of
.450 or higher. Yet no major league baseball player has ever
finished the season with a better than .450 average. What do you
think is the most likely explanation for the fact that batting averages
are higher early in the season?
• One time at bat has a much greater effect on one’s average early in
the season than at the end. For example, if someone bats twice
after two weeks and gets one hit, his average is .500, but it may not
be a true indication of how well he bats. The more frequently he
bats, the clearer the true information as to how well a batter hits.
 The answer represents an understanding that averages based on
more cases are more reliable (that is, less variable) than averages
based on but a few cases.
Samples Critical Thinking
•
David L., a senior in HS on the East Coast, was planning to go to college. He had
compiled an excellent record in HS and had been admitted to his two top choices: a
small liberal arts college and an Ivy League university. The two schools were about
equal in prestige and were equally costly. Both were located in attractive cities, about
equally distant from his hometown. David had several older friends who were
attending both schools. They were all excellent students like himself and had
interests similar to his. The friends at the liberal arts college all reported that they
liked the place very much and that they found it very stimulating. The friends at the
Ivy League university reported that they had many complaints on both personal and
social grounds and on educational grounds. David initially thought he would go to the
smaller college. However, he decided to visit both schools himself for a day. He did
not like what he saw at the private liberal arts college. The people he met seemed
cold and unpleasant; a professor seemed abrupt and uninterested in him. However,
he did like what he saw at the Ivy League university. The people he met seemed
vital, enthusiastic, and pleasant. The two professors he met took a personal interest
in him and he came away with a very pleasant feeling about the campus. Please say
which school you think David should go to and why.
• David just saw the Ivy League university for one day. His friends’ reports are based
on an entire year. So he should take his friends’ word for it. Chances are that the
liberal arts college is better.
 The answer represents an understanding that averages based on more cases are
more reliable (that is, less variable) than averages based on but a few cases.
Research Design
Sample
Collection of subjects (Ss)
used in a study, part of
the population
Random Sample – everyone in the population
being studied has an equal chance of being included
in the study
Stratified Sample – identified
subgroups in the population
are represented proportionately
in the sample. EX: 12% of the
American population is African
American. A stratified sample
would thus be 12% African
American.
The larger the sample, the more likely
it will represent a cross section of the entire population
Sample
Population
Larger collection of people about which we want
to generalize
Population
Sampling Bias
When the sample is not
representative of the larger
population.
EX: Study on voters – survey
done over the phone. Biased
sample b/c lower economic
groups may not own telephones
Research Methods
5.
Carry out Observation
 Descriptive Methods (naturalistic
observations, case studies, surveys) –
used it to collect and describe data
 Correlational Study – use it to
reveal how closely two things vary
together and thus how well the
presence of one variable predicts the
presence of another variable.
 Key Point: Correlational study
cannot prove cause and effect!!!
The relationships between the
two variables can be a result of
many factors
 If relationship is strong enough,
than….
 Experiment – manipulate
independent and dependent
variable to prove cause and effect
Analyze the Data + Report Findings
6.
7.
Analyze the data and draw
conclusions. Based on results you
might have to refine theory.
 Report findings precisely enough
to permit replication and
revision of theory
 Replication = the
experiment can be repeated
and would yield constant
results even when done with
a different group of people
or by a different person
Depending on results, start back at
step #1
Types of Descriptive Studies
Type of
Study
Description
Advantages
Disadvantages
Case Study
An observation technique in
which one person, group, or
situation is studied in depth in
the hope of revealing
universal principles (often
combines observations,
interviews, tests, and
analyses of written records)
Useful in studying rare
or complex phenomena
Can mislead us - any given individual
may be atypical and lead to false
conclusions
Survey
Questionnaires or special
interviews administered to a
large, random group of
people to ascertain their selfreported attitudes or
behaviors that cannot be
directly observed
Enable researchers to
describe the
characteristics of a
relatively small sample
(a few hundred people)
and then generalize
that information to a
larger population,
quickly collect
Easily biased (tend to hang around
people like us)
A small return rate means that the
sample is not representative
Wording of questions can have major
effects on the opinions respondents
express (framing – the way an issue is
posed can significantly affect decisions
and judgments)
Not in-depth, not answered truthfully
Naturalistic
Observation
Method of gathering
descriptive information,
involves watching behaviors
of interest, without interfering,
as they occur in their natural
environments.
Obtain data about a
truly natural behavior
rather than a behavior
that is in reaction to
contrived experimental
situation
If participants realize they are being
observed, their behavior becomes
unnatural
Difficult and time-consuming
Controls are lacking
Difficult to generalize the results of the
research
Correlation Coefficient = r
•
Correlational Study – use it to reveal how closely two things vary together and thus
how well the presence of one variable predicts the presence of another variable.
– Key Point: Correlational study cannot prove cause and effect!!! The
relationships between the two variables can be a result of many factors
– Positive correlation - Two variables change in the same direction, as x increases
so does y. (max +1.00)
– Negative correlation - As one variable goes up, the other goes down and visa
versa. Inverse relationship, as x increases y decreases. (min –1.00)
– Or not related! (~ 0.0)
• The higher the absolute value of r, the stronger the relationship
• Perfect correlation r = + or – 1.00
• EX: r = +.37
– “+” r “-” indicates direction of relationship
– .37 indicates strength of relationship (0.00 to 1.00)
Perfect positive
correlation (+1.00)
No relationship (0.00)
Perfect negative
correlation (-1.00)
Correlation Coefficient
Indicates direction
of relationship
(+ or -)
Correlation
coefficient
r = +0.65
Indicates strength
of relationship
(0.00 to 1.00)
Correlation CANNOT Prove Causation
• Three Possible Cause-Effect Relationships
(1)
High self-esteem
could cause
High achievement
or
(2)
could cause
High achievement
High self-esteem
or
High self-esteem
(3)
Supportive parents
or biological
predisposition
could cause
and
High achievement
Experimental Variables
Independent Variable (IV)
Condition or Event that the experimenter
varies. Hypothesized to cause an effect on another variable. Given
to experimental group
Dependent Variable (DV)
Variable thought to be affected by the IV.
What you measure.
Alcohol
Level
Performance
Experiments
Random
Assignment
Control
Condition
Everything the same as
experimental group,
except don’t get IV.
Maybe get placebo
Experimental
Condition
IV
Random assignment - every subject in the study should have an equal
chance of being placed in either the experimental or control group (even/odd;
draw straws)
 Minimizes pre-existing differences between those assigned to the
different groups (age, attitudes, etc). Any later differences between people
in the experimental or control conditions must be the result of the
treatment.
Sources of Bias
• Observer-expectancy effect (Experimenter Bias)
– researcher has expectations that influence measurements. EX: send
subtle nonverbal signals, make mistakes in recording subjects’
responses.
• Subject-expectancy effect
– subject knows design and tries to produce expected result.
– demand characteristics – any cues in a study that suggest to
subjects the purpose of the study or what the researcher hopes to find.
• Volunteer bias
– People who offer or volunteer to participate in research studies differ
from people who do not (more interested in research than nonvolunteers, have more spare time, more willing to disclose intimate
information, etc)
• Placebo effects
– Placebo is a physical or psychological treatment that contains no active
ingredient but produces an effect on the dependent variable because
the person receiving it believes it will.
• Prevent bias by blinding!!
– minimize expectancy by removing knowledge about experimental
conditions
Experimenter Flaws
Extraneous Variable
Anything other than the IV that may influence
the DV. Differences in subject variables (age, gender, race,
ethnicity, cultural background, socioeconomic status, IQ,
health, etc) and situation-relevant variables (test conditions,
experimenter behavior, timing, etc)
Confounding of Variables
When two variables (an extraneous variable and IV) are
linked such that it is difficult to sort
out their specific effects on the DV
Prevent it by Random Assignment!!
Good experiment can be replicated – the experiment
can be repeated with a different group of people
and would yield constant results
Experimental Design: Placebo
• Placebo Effects –
placebo is a physical
or psychological
treatment that
contains no active
ingredient but
produces an effect on
the dependent
variable because the
person receiving it
believes it will.
Review
1. What is the purpose of correlational research?
2. The correlation between the physical weight and the reading ability of
elementary school students is +.65. What does this mean?
3. If those who watch a lot of TV violence are also particularly likely to behave
aggressively, this would NOT necessarily indicate that watching TV violence
influences aggressive behavior. Why?
4. Which is a stronger correlational relationship, +.34 or -.85?
5. Karen dreamed that a handsome young man she had met the previous day
asked her for a date. When he actually did call for a date several days later,
Karen concluded that dreams accurately predict future events. Her belief
best illustrates what?
Ethical Guidelines in Psychological Research
•
Psych departments have review panels
(Institutional Review Board) to screen all
proposals for research. Proposals have
to conform to APA Ethics Code
Accurately report results
•
•
Right to privacy - studies are conducted to
ensure confidentiality
– information will not be made available
to anyone who is not directly involved
in the study
– data is coded and reported in group
form rather than individual responses.
Use code #s instead of individual
names whenever possible.
Informed consent
– use of deception – not possible or
desirable to always tell the truth about
purpose of experiment up front (want
to eliminate subject expectancy effect)
– always have to debrief subjects after
study – subjects should be told about
the purpose of the research, the
hypotheses being tested, the nature of
the results or anticipated results, and
the implications of those results for the
science of psychology immediately or
soon after participation.
Ethical Guidelines in Psychological Research
• Minimize subject discomfort
– subjects can withdraw any time
and have to sign consent
• Prevent any long-term negative
effects
– Subjects should not be put at risk,
except in cases where there is no
other way to conduct research
and research results offer
promise to advance knowledge
– Voluntary Participation - people
cannot be coerced into
participating in research.
• Right to Service - treatment might
have beneficial effects, so persons in
no-treatment control group may be
upset
Animal Experimentation
•
Review boards (Animal Care and Use
Committee) establish Animal Care
Guidelines:
– Specify the proper care and
maintenance of animals, including the
adequacy of space requirements,
maintenance of good health, and
supervision by qualified personnel.
– Researchers are required to make
every effort to minimize the pain and
discomfort of animal subjects and to
seek alternative methods for their
research if possible.
– Similar processes often underlie
animal and human behavior (how
humans see, exhibit emotion, become
obese, etc), so the justification for
discomfort or harm a research
procedure may produce is that the
results will be applicable to humans
Tom the Dancing Bug
• Cartoon reinforces the point that we value animals
according to their perceived similarity to us.
– Speciesism – prejudice toward the interests of one’s own
species and against the interests of another species.
• Arguments against research using animals:
– Just as differences in intelligence, race, and gender are not valid
criteria to exploit other humans, a creature’s species is equally
irrelevant
– All sentient animals have the capacity to suffer, and thus are the
subject of equal moral consideration.
– Research with animals is permissible only if we would also
consider using human subjects for the same experiments
– Like humans, animals have the right to be treated with respect
and the right not to be harmed.
Animal Research
Proposals: Case 1
• Approve or Reject? Why?
• Forces consideration of whether injury to
another species closely related to human is
justified if the results will be applicable to human
beings.
• Review Board would most likely say YES 
similar processes often underlie animal and
human behavior, so the justification for
discomfort or harm a research procedure may
produce is considered acceptable if the results
will be applicable to humans
Animal Research
Proposals: Case 2
• Approve or Reject? Why?
• Prompts consideration about the use of
animals when there is no direct human
application
• Review Board would most likely say YES
 pure research in scientific progress is
important. Need to make sure animal
discomfort is minimized and no other
method of research is available
Animal Research
Proposals: Case 3
• Approve or Reject? Why?
• Involves the question of whether pound
animals should be used in research
• Review Board would most likely say NO
many states have banned the use of
such animals for biomedical research of
for student surgeries in veterinary schools
Animal Research
Proposals: Case 4
• Approve or Reject? Why?
• Prompts consideration about using
animals in student laboratories.
• Review Board would most likely say NO
animal welfare groups argue that this is
particularly unnecessary when videotapes
and computer simulations are adequate
substitutes
Statistics
Use of mathematics to organize, summarize and interpret numerical data.
Statistical analysis is used to determine whether any relationships
or differences among the variables are significant, quantifies the exact
strength of the association.
Descriptive Statistics
Used to describe, organize
& summarize data to
make it more understandable
Statistical Significance
Central Tendency
Variability
Correlation
Used to interpret data
& draw conclusions. “What can we infer
about the pop from data gathered
from the sample?”
Inferential Statistics
Descriptive Statistics: Measures of Central Tendency
(summarizes data set by providing a representative number)
Median
Score that falls in the center of a distribution of scores.
When there is an even number of scores in a data set, the
median is halfway between the two middle numbers.
Best indicator of central tendency when there is a skew.
The median is unaffected by extreme scores.
Mean ∑ X/N = X
Average of scores in a distribution. Even one
extreme score can change the mean radically,
possibly making it less representative of the
data. Most significant because additional
statistical manipulations can be performed on
it.
Mode
Most frequently occurring score in a distribution.
Skewed Distribution
•
An asymmetrical distribution of scores, such as a curve
with a bump on the left and tail to the right or most
scores are bunched to the left or right of the mean
– The mean is the largest
– The mode or median are smaller than the mean
– The mean is a less useful measure; while the median is
more useful
15 20 25 30 35 40 45 50
90
475
70
Mode Median
One Family
Mean
Income per family in thousands of dollars
710
Descriptive Statistics: Measures of Variability
Indicate the dispersion or spread in a data set. How much the scores
in a set of data vary from:
a. Each Other
b. the Mean
Tell you if the scores are very different from one another or if they
cluster around the mean.
Range
The difference between the highest and lowest score in a set of data.
Extreme scores can radically affect the range of a data set.
Standard Deviation
Reflects the average distance between every score and the mean. Tell
You how different the scores are from the
mean. Tells you whether scores are
packed together or dispersed.
Variability
Standard
Deviation
Easy Example of Standard Deviation
•
•
Memorize formula
– X = mean
– X = individual score in data set
– ∑ = sum/add
– N = number of scores in data set
Sample Problem
– Find the standard deviation of the following data set (2, 4, 4, 6)
1. Find the mean of data set: (2+4+4 +6)/4 = 4 = X
2. Subtract each individual score in data set from mean (X – X) and
square it
• (2 – 4) = (-2) = 4
• (4 – 4) = (0) = 0
• (4 – 4) = (0) = 0
• (6 – 4) = (2) = 4
3. Add each squared score: (4+0+0+4) = 8
4. Divide summed scores by N (which is 3 in this example, since there
are 4 scores in the data set): 8/2
5. SD = 2
Inferential Statistics
• While descriptive statistics summarize a data set, we
often want to go beyond the data:
– Is the world at large like my sample?
– Are my descriptive statistics misleading?
• Inferential statistics give probability that the sample is
like the world at large.
– Allow psychologists to infer what the data mean.
– Assess how likely it is that group differences or
correlations would exist in the population rather than
occurring only due to variables associated with the
chosen sample.
Statistical Significance
Results are “statistically significant” when the probability
that the findings are due to chance is very low.
If the difference between two group means is statistically
significant, a researcher would conclude that the difference most
likely exists in the population of interest. If the difference is not
statistically significant, a researcher would conclude that the
difference occurred by chance – possibly because of an
unrepresentative sample or the presence of confounding variables.
“Very Low” means less than 5 chances in 100
or
P < 0.05 level of significance
68.2 %
95.4 %
99.7 %
Sample Test Question
• For a language test with normally
distributed scores, the mean was 70 and
the standard deviation was 10.
– How are the scores distributed?
– Approximately what percentage of test takers
scored 60 and above?
Answer to Sample Question
• 68% of students scored between 60 and
80.
• 84% of students scored 60 and above
Educational Cartoons + Intelligence
•
Correlational Study
– Population and sample? How collected?
- All preschoolers enrolled in preschools in Deerfield.
- Randomly sample 100 preschoolers from the population (enter names in
computer, and computer randomly generates 100 names).
– How will you collect the data?
• Send a survey home to the parents of the 100 preschoolers asking them
to report on how many hours of educational cartoons their child watches.
• Review the preschooler’s scores on their entrance exam to the preschool
• Compare the data from the parents’ surveys and the scores from the
entrance exam by creating a scatterplot or determining the correlation
coefficient.
– Variables? Operational Definitions?
• Hours of educational cartoons – Baby Einstein videos
• Scores from the entrance exam – a rating of above average or exemplary
would indicate a high degree of intelligence
– Ethical Guidelines?
• Informed consent from parents, confidentiality, debriefing
•
Correlational Study
– Meaning of Results: The correlation coefficient “r” = .83. Should Philip
continue his research and design an experiment? Why or why not?
• Yes! There is a strong, positive relationship between educational
cartoons and intelligence. However, we do not know which
variable influences the other variable. It could be…
(1)
Educational
cartoons
(2)
Smart children
could cause
Increase in IQ
or
prefer
or
(3)
Socioeconomic
background of
parents
have enough $
Watching
educational
cartoons
Buy educational
cartoons
and
Hire tutors to
tutor their
children
Educational Cartoons + Intelligence
•
•
•
•
Experiment
Population and sample? How collected?
- All preschoolers enrolled in preschools in Deerfield.
- Randomly sample 100 preschoolers from the population (enter names in
computer, and computer randomly generates 100 names).
Variables? IV/DV? Operational Definitions?
– IV = educational vs non-educational cartoons
• Educational cartoons = 1 hour of Baby Einstein’s every day for a month
• Non-educational cartoons = 1 hour of Sponge Bob every day for a month
– DV = intelligence
• scores on an IQ test
Experimental and Control Group? How are subjects assigned?
– Experimental group =
• Group 1 of preschoolers that watch 1 hour of BE every day for a month
• Group 2 of preschoolers that watch 1 hour of Sponge Bob every day for a month
•
– Control group = Normal average daily T.V. Watching.
– Randomly assign 50 preschoolers to experimental group (2 Variables being given) and
50 to control group.
Confounding Variables or Bias? Ways to Control?
–
–
–
Double Blind to prevent bias and subject-expectancy effect
Differences in intelligence to start with – prevented by random assignment.
Differences in family involvement – prevented by random assignment
Educational Cartoons + Intelligence
• Ethical Guidelines?
– Protect confidentiality – data is coded
– Parents need to give consent
– Parents need to be debriefed after study is over
– Minimize subject discomfort and prevent negative long-term
effects
• Meaning of Results:
– The mean of the experimental group was 83 and the mean of the
control group was 79. If a statistical test finds that the statistical
significance “p” < .05 for the difference between the means, what
should Philip conclude? Why?
• Philip can conclude that there is less than 5 in 100 chances
that the results were due to chance. Most likely the
treatment (IV – educational cartoon) created the difference
between the two groups.
Statistical Breakdown of Intelligence
• After researching about the intelligence of preschoolers, Philip was curious to
find out how intelligent he and his friends were. He and his friends took an
IQ test. The scores for the IQ test were normally distributed – the mean was
100 and the standard deviation was 15. Using this information, describe how
the scores are distributed.
–
Most scores (68%) are within 15 points of the mean (of 100).
–
The typical (average, normal) IQ score falls between 85 and 115
–
Mean, median, and mode are the same or very close
Statistical Breakdown of Intelligence
• The following are a list of four scores from the IQ test Philip’s
friends took: 136, 95, 91, 90. If we wanted to know what the
IQ of Philip’s friends is MOST like, which would be the best
indicator? Mean or Median? Why?
– X = 103 Median = 93
– Answer = Median. The mean is affected by extreme
scores.
Statistical Breakdown of Intelligence
• Philip then wanted to find out if he and his friends were smarter than
his dad and his dad’s friends, so he gave the IQ test to his dad and
his friends. Compare the two groups of scores:
– Philip’s group: 104, 102, 95, 91, 90, 83, 72
– Philip’s dad’s group: 95, 93, 92, 91, 90, 89, 87
• What can we determine about the two groups? How are they
different? Similar?
– For each of the groups, the mean = 91
– However, the range for Philip’s group = 32; while the
range for dad’s group = 8. The standard deviation for
Philip’s group = 10.23; while the standard deviation for
Dad’s group = 2.45
– The groups did not perform the same. The scores in
Philip’s group are much more spread out than in the
dad’s group. The scores for the dad’s group tend to
cluster closer to the mean
Statistical Breakdown of Intelligence
• The IQ test Philip used was recently re-normed. Why are IQ
tests periodically updated?
– Changes in knowledge require tests to be renormed
• People have gotten smarter (Flynn Effect)
• The numbers of questions answered accurately
has increased over the years
• Changes that affect IQ test scores of groups (e.g.
sociocultural or technological)
• Changes in educational practices or techniques
(that affect knowledge)
• Keep material culturally relevant
– Re-norm to maintain validity or reliability
Flynn Effect
In the past 60 years, intelligence scores have risen steadily
by an average of 27 points.
The following environmental changes have contributed to the change:
1.
Rise of science:
a.
Taught us that classifying the world using the categories of science is just as
important as manipulating the world
b.
Freed logic from the concrete, allowing us to work on abstractions with no
concrete referents.
2.
Increasing educational opportunities
3.
Reduction in family size
4.
Improvements in infant nutrition
5.
Changing communication technologies
Principles of Test Construction
For a psychological test to be acceptable it must fulfill the
following three criteria:
1.
Standardization
 Standardizing a test involves pre-testing a
representative sample of people and forming a
normal distribution or bell curve (most scores fall
near the average, and fewer and fewer scores lie near the
extremes) to establish a basis for meaningful
comparison.
2. Reliability = a test is reliable when it yields consistent
results
3. Validity = a test is valid when it measures what it is
designed to measure.
Reliability
To establish reliability researchers establish
different procedures:
1.
2.
3.
4.
Split-half Reliability: Dividing the test into two equal halves
(odds and evens) and assessing how consistent the scores are.
Alternate Forms Reliability: Using different forms of the test
to measure consistency between them.
Test-Retest Reliability: Using the same test on two occasions to
measure consistency. A person’s score on a test at one point in
time should be similar to the score obtained by same person
on a similar test at a later point in time.
Inter-Scoring Reliability: One scorer’s rating should be similar
to another scorer’s rating
Validity
Reliability of a test does not ensure validity. Validity of a test
refers to what the test is supposed to measure or predict.
1.
Content Validity: Refers to the extent a test
measures a particular behavior or trait of
interest.
 driving test that samples driving tasks
2.
Predictive (Criterion-Related) Validity: Refers to the
success of a test in predicting a particular behavior or
trait it is designed to predict. Assessed by computing
the correlation between test scores and the criterion
behavior.
 behavior (such as college grades) that a test (such
as the SAT) is designed to predict
Types of Tests: Aptitude vs. Achievement
 Aptitude = A test designed to
predict a person’s ability to
learn a new skill (future
performance.)
 College entrance exams
like ACT and SAT
 http://www.pbs.org/wgbh/p
ages/frontline/teach/diversit
y/sat/sat.html
 Achievement = A test
designed to reflect what you
have already learned.
 Unit exams
 AP exam