• based in limited information available from the observed event
• that is what our previous experience has told us

• based on evidence of the senses
• must be based on objective and systematic obserations
• often carried out under carefully controlled conditions
• must be verified by others

verify through direct obervation or lead to specific predictions about what should occur under conditions not yet observed

can maintain confidence in an explanation but are always willing to entertain the possibility that the explanation is faulty

• empirical
• general
• parsmonious
• rational
• rigorously tested
• tentative
• testable

• explanations that are derived from trusted sources
• no evidence is required
• tend to be un-parsimonious, are assumed to be true, lack generality and fail to produce testable predictions

• observe a phenomenon
• formulate tentative explanations
• further observing and experimenting
• refine and retesting

step 1 of the method

• observes the behavior of interest
• observes the phenomenon in many different ways
• identify the variable of interest
• observe influence on behavior

step 2 of scientific method

• develop one or more tentative explanations that seem consistnent with your observations
• tentative explanations are called hypothesis

step 3 of the scientific method

devise a study or studies that test your explanations

step 4 of the scientific method

• generate new, more specific hypotheses in light of previous results
• create new studies using your new hypohesis

•  
  •  
    • scientific,
    • common sense,
    • belief based

basic

applied

• focus on on confirming or disconfirming theories
• goal to acquire general information about a phenomenon and either validate or invalidate previous theories
• little emphasis on applicaitons to real-world examples
• focus is to contribute to the theories of human behavior

• focus on investigaing a real world problem
• the goal of its research is to generate information that can be applied to the solution of a real-world problem
• focus is to obtain answers to pressing questions about reality

1. Descriptive
2. Relational
3. Causal

• describes what is going on and what exisits
•  focused on describing some phenomenon

1. develop a research idea and hyphothesis
2. choose the research design
3. choose participants
4. decide what to observe and appropriate measures
5. conduct the study
6. analyze the results
7. report the results
8. redo the process

Step 1

• identify the issue you want to study
• develop a research question that can be tested empirically
• state the hypothesis

step 2 of the research design

• what type of study (correlational, experimental, quasi-experimental)
• where you are going to conduct your study
• how you are going to measure the behavior of interest
• sometimes pilot study first

step 3 of the research design

• what type of participants you are going to use
• how you are going to treat your participants
• ethics

step 4 of the research design

• decide on what behaviors to observe 
• how you are going measure behaviors

step 5 of the research design

• participants take part in the study
• observe and measure their behavior
• record all data

step 6 of the research design

• summarize and analyze data,
• perform descriptive and inferential statistics

step 7 of the research process

• format paper using apa
• present or submit

step 8 of the research process

• use this study to ask questions for further study

• variable manipulated by the research
• what the researcher is testing to see where the differences lie
• can be predefined groups

• variable the researcher is trying to measure
• behavior of the phenomenon depends on the behavior of the participant

• variables that may affect the behavior that you want to investigate but you are not interested in
• to counteract either control or assign participants randomly

• this is the research hypothesis
• contains the variable that is the prediction of the study

fail to reject a false null hypothsis says that there is no difference when one truly does exisit.

 decrease alpha ie from .05 to .01 then increase the chance of making a type II error

• defines the probability of making a type I error
• designated as alpha level established a piori (before begin research)
• states how much risk you are willing to take for type I error
• the most common alpha levels in psychology are .05 and .01
• decrease alpha then decrease the chance of a type I error

• How well you can detect a difference,
• the probability of rejecting a null hypothesis
• one tailed tests are more powerful than 2 tailed
• decrease alpha then you decrease power
• increase sample size increase power
• increase effect size then you increase power

• developed after the nazi trials
• has 10 points researchers are to follow

• developed 1964 by world medical assoc
• addresses medical research several principles

developed 1979 and has three main points,

respect for persons, beneficence, justice

first developed in 1953

latest revision 2002

• informed consent
• debriefing
• minimal risk

• oral or written consent by participants before research
• volunteer info about the study what will the participant do, risks, and/or benefits
• the right to refuse and withdraw at any time

• explain after the research in some detail given the opportunity to ask questions
• written or oral

• research should minimize risks
• two major risks invasion of privacy and stress

• alter data
• plagerize
• used unethical practices
• may commit for personal, tenure, or financial reasons
• may or may not be reported

• avoid duplication of research
• reveal other questions of topics of interest
• provides with reasons for addressing various design questions
• keeps you up to date on research topic of interest

• scholarly
• substansive news/general interest
• popular
• sensational

• reports summary of research
• does not provide all details to duplicate

• books
• journals
• conventions/professional meetings
• personal communication
• internet

• look for relationship between two variables
• not interested in cause and effect
• do not manipulate

• can study variables you cant manipulate
• it can be applied to more real world problems
• more generalizable than experimental

• find differences among groups
• control over variables
• manipulate the independent variable and control extraneous
• can establish cause and effect

• can only be used in certain situations
• can control too many extraneous variables that research is artificial and difficult to generalize

• differences among groups
• do not manipulate the independent variable
• cannot assign participants to groups

• set of categories that have different names
• discrete or categorical variable
• least precise

• qualitative technique
• set of categories
• ordered sequence observations ranked
• intervals between categories are not equal

• quantitative data
• ordered set of categories
• categories form a series of intervals that are the same size
• no absolute 0
• continuous variable

• quantitative measure
• same as interval scale plus absolute 0
• most precise continuous variable

• nominal
• ordinal
• intervial
• ratio

• nominal - least
• ordinal  - some quantity
• interval - how the values differ
• ratio- most information

parameteric stastics (t-tests, anovas) require interval or ratio data

nominal or ordinal - nonparameteric tests -

chi square, mann whitney U test

• strength of conclusions, inference, or proposisitons
• study testing what it is supposed to be testing

• is the study measuring what it is supposed to be measuring
• experimental research have higher levels than quasi experimental

• history
• maturation
• testing
• instrumentation
• statistical regression
• mortality
• selection threats

• are results representative and generalizable
•  questions what populations, settings, treatment variables and measurement variable can this effect be generalized
• experimental research has lower levels where quasi experimental higher

• interaction of testing and treatment
• interaction of selection biases and the experimental variable
• reactive effects of experimental arrangments
• multiple treatment inference

• does measure produce similar results when repeated under identical conditions measure consistent
• should be at least .80

test-retest

parallel forms

split-half

internal consistency

inter-rater or inter-observer

reliability where you give the same measure more than one time to the same group

reliability where you give a measure to a group then give a similar measure at a later time

• how well the items in the measure yield the same results -
• consistnent results are for different items on the same measure
• cronbach's alpha

• assess the reliability of raters or observers in a study
• raters/observers are measuring instances of behavior in subjects

extent to which your measure measure what it is intended ot measure

• construct
• face
• content
• predictive
• convergent
• concurrent
• discriminant

• assess how you have defined the variables in your measure on its face --
• seems like a good measure
• weakest

• how well the questions in your measure compare to other questions

assess how well your measure relates to another measure of the same construct

how well your measure is able to discriminate between groups that it should theoretically be able to distinguish between

• chance variation among samples selected from the same poputlation
• select a large enough sample to reduce sampling error

• error that is introduced by a sampling procedure that favors certain characteristics over others
• find source of bias and reduce it
• increased sample size doesnt help

• uses some form of random selection
• assures each participant has the equal chance of being selected for your study

• Simple random sampling
• stratified random sampling

• non-random likelihood of selection is not known
• not everyone in the population has an equal chance of being selected

• Puposive sampling (there are 5 types)
• Accidental or convenience

individuals thought to be most important or relevant to the issue studied are targeted for the research

• snowball-- ask participants for names of others with same characteristic
• quota - select participants according to some fixed quota
• expert- get an sample from experts in an area
• modal instance - sample the most frequent case
• heterogeneity sampling - get the most diverse sample you can get