Week 3

Question 1

Scientific measurement

Answer 1

Our measurement of the dependent variable are the basis for evaluating our hypotheses and should meet the following standard:
- Objectivity
- Reliability
- Validity
- Sensitivity

Question 2

Scientific Measurement: Objectivity

Answer 2

Objective measurements minimize the influence of bias and individual interpretation.
- Free from subjective preferences or attitudes.
- Some techniques (clinical exams, field studies, interviews) rely on the skill and experience of individuals and are more susceptible to bias.

Question 3

All observations are theory-laden

Answer 3

They are influenced by the meaning of terms, the n azure of the tools used, and the theoretical context.

Question 4

Scientific Measurement: Reliability

Answer 4

Realizable measurement should give consistent results each time and across experimenters.
- Applies to instruments and procedures.
- Ex: fMRI is widely used technique, but the measurement depends on software used, scanning methods, and statistical analyses. Different groups doesn’t always get the same results.

Question 5

Scientific measurement: Validity

Answer 5

Validity questions the extend to which a technique or experiment measures what it intends to measure.

Question 6

Internal Validity

Answer 6

The design of the experiment should ensure that the independent variable (and not some confound) is responsible for the change in the dependent variable.

One threat is task demand characteristics: Something in the task (e.g. wording of a question) or environment may indicate what the participant is supposed to do.

Question 7

Experimenter effects

Answer 7

Experimenters may unintentionally affect the outcome of an experiment, reducing internal validity.

• They expectations may affect how they deliver instructions, record results, interact with participant, etc.
• To limit this, experiments often use automated procedures.
• You can also use blinded design (experiment and or subject is unaware of the condition)

Question 8

External Validity

Answer 8

How does your result generalize beyond a specific experiment.
- Sampling bias
- Extraneous variables

Question 9

Sampling bias

Answer 9

The participants in the current experiment may be different from the general population.

Question 10

Extraneous variables

Answer 10

Some factor (e.g. temperature) may not be a confound in the current experiment, but may change results in other experiments.

Question 11

Population

Answer 11

A complete collection of observations (data) or potential observations for all individuals or units of interest.

• This is defined by the investigator. The population can be targeted (e.g. heat tolerance in men over 65) or very broad (e.g. pain tolerance in all adults).

Question 12

Sample

Answer 12

A partial set of observations taken from the population.

• Typically much smaller than the population (e.g. 20 volunteers from a community center).
• Should be representative of the population of interest.
• However, experimenters use a convenience sample in that it is not randomly drawn from the population of interest (e.g. SONA).

Question 13

WEIRD

Answer 13

Western, educated, industrial, rich, and democratic.

Question 14

WEIRD samples

Answer 14

Basic reasoning and judgements of fairness vary widely across different cultures.

Question 15

A hypothesis is a prediction that can be expressed in two forms:

Answer 15

1. Research hypothesis
2. Statistical hypothesis

Question 16

Research hypothesis

Answer 16

Something posed as a question, suggesting a specific relationship between two variables.

Does caffeine improve declarative memory?/Consuming famine will increase the number of items recalled from a list.

Question 17

Statistical hypothesis

Answer 17

Proposes two competing statistical values, only one of which can accurately express the nature of the relationship between two variables.

1) After drinking 100 mg of caffeine, the avg. number of words from a list recalled > avg. number of words from a list recalled without caffeine.

2) After drinking 100 mg of caffeine, the avg. number of words from a list recalled <= avg. number of words from a list recalled without caffeine.

Question 18

_______ begin with the assumption that no special relationship exists between your variables.

Answer 18

Statistical hypothesis.

Question 19

Bull hypothesis H0

Answer 19

A statistical hypothesis asserting that our stated research effect will not be found. (Nothing special is happening)

-GPA of ‘more study’ samples <= GPA of ‘’less study’ sample.

Question 20

Alternate hypothesis H1

Answer 20

A statistical hypothesis asserting the opposite of the null. (Something special is happening.) This agrees with your research hypothesis.

• GPA of ‘more study’ sample > GPA of ‘less study’’ sample.

Question 21

If the experiment demonstrates your predicted relationship between the
variables, you reject the ____ hypothesis. If not, you fail to reject (or you
retain) the ____ hypothesis. The result is _____ _______.

Answer 21

Null; null; never proven.

Question 22

Statistical hypothesis (null and alternate) must be ____ ____ and _____ _______.

Answer 22

Mutually exclusive; Together exhaustive.

Question 23

Mutually exclusive

Answer 23

your results (measured values) must either be consistent with the null hypothesis or the alternate hypothesis – it can’t support or falsify both.

• Studying more can’t both increase and decrease GPA

Question 24

Together exhaustive

Answer 24

All results must fit into one of your hypotheses. The null and alternate cover 100% of all possible measured values.

• All possible results (i.e. measured values of dependent variable) must be included in the statistical hypotheses.

Question 25

When we fail to reject the null hypothesis (H0), do we prove the null is true?

Answer 25

No - Null hypothesis was likely or unlikely to explain the measurements taken. - The null hypothesis can never be proven.

Question 26

There are ____ ways to reject the null

Answer 26

Infinite

Question 27

Why do we test the null hypothesis and not the research (alternate) hypothesis?

Answer 27

- The research hypothesis lacks the necessary **precision.** - The decision to reject the null hypothesis is **stronger decision** than failing to reject. Weak decision if you fail to reject and stronger decision if you reject the null, supported by an improbable (rare) results.

Question 28

In any experiment, there are two possible states of the world.

Answer 28

1. H0 is true 2. H1 is true These are mutually exclusive and together exhaustive (probabilities sum to 1)

Question 30

Type I Error

Answer 30

This is the error when making a decision to reject the null hypothesis when it is actually true in the population. - no effect is present - We call it false alarm/positive because the decision says a real effect exists when it does not. - chosen by the experimenter. P=a(alpha)

Question 31

Type II Erroor

Answer 31

This is an error when the decision to retain the null hypothesis is wrong because the null is false in the population. (And the alternate is true) - an effect is present - miss/false negative - probability of a miss is p=B (beta)

Question 32

If you retained the null hypothesis, which of the following errors could you have made? A. Type I Error B. Type II Error C. Both D. Neither

Answer 32

B. Type 2