Research Psychology exam 4

2 common reasons to summarize data

To clarify patterns observed & conciseness

Before summarizing data

Identify possible omissions, errors, or other anomalies. This removes all errors before further analysis.

How to analyze your data set

Describe it and determine how to interpret it

Measures of Central Tendency

Mean, Median, and Mode

Frequency

How often a value occurs in the form of tables or graphs

Frequency distribution table

Summarizes distribution of the data in terms of how of how often the value occurs. It can be continuous, categorical, or discrete. The type of data determines the data category there’s individual and interval scores.

Types of graphs for data types

continuous data: histogram and discrete data: bar chart or pie chart. Pie charts are usually used when there’s more categories than bar charts.

Mean

The sum of all the scores divided by the number of scores. It is the average. It describes normally distributed data with an interval or ratio scale.

Median

The midpoint of the data distribution. Use the median when data is skewed that way that data isn’t intensely impacted by the mean. An example is if we calculated our annual income and included Beyonce. If she joins then our mean income would skyrocket and not be indicative of the population. Ordinal scale.

Mode

Value that occurs the most often. There may be one or more. Used in biology or in retail/small businesses. Nominal scale.

Deviations from the normal scale

One factor is kurtosis or how peaked the distribution is. Uniform distribution is when all scores have the same frequency. Uniform distribution is when there’s two distinct peaks. Skewed distribution is when most scores cluster at one end of the distribution.

Mesokurtic

Kurtosis with a moderate peak

Leptokurtic

Kurtosis with a high peak (most scores are clustered in the middle)

Platykurtic

Kurtosis with a flat peak (scores are more spread out)

Variability

Dispersion or spread of scores in a distribution

Range

Difference between the largest and smallest value. Best for data sets with outliers.

Interquartile range

divides the distribution into quarters.

Variance

Average squared distance of sample scores from the mean

Degrees of freedom

The sample size minus one

Sum of squares (SSS)

Sum of squared deviations from the mean

Standard deviation

Average distance scores deviate from the mean. Always report standard deviation when you report the mean.

Empirical rule

Rule for normally distributed data at least 99.7% of data falls within 3SD, 95% falls within 2SD, 68% falls within 1SD.

Types of graphs

Bar graphs for nominal or ordinal scale data, Line graph for interval or ratio scale data, and scatterplot which is a graphical display of data points

Between subjects design

Different participants are observed one time in each group or at each level of a factor (everyone gets treatment A). It is the only design that can meet all three requirements of an experiment a.k.a randomization, manipulation, and inclusion of a comparison/control group but the sample size needed can be large.

Between subjects experimental design

Levels of a between subjects factor are manipulated then different participants are randomly assigned to each group or to each level of that factor and observed one time. (Everyone gets treatment A)

Control

The manipulation of a variable while holding all other variables constant

Experimental or treatment group

Participants are exposed to a manipulation (IV) that is believed to cause a chance in the dependent variable

Control Group

Participants treated the same as those in an experimental group except that the manipulation is omitted

Experimental manipulation

Identification of an IV and the creation of two or more groups that constitute the levels of that variable.

Natural manipulation

Manipulation of a stimulus that can be naturally changed with little effort

Staged manipulation

Manipulation that requires the participant to be “Set up” to experience stimulus or event.

Random Assignment

Procedure used to ensure that each participant has the same likelihood offing selected to a given group

Restricted random assignment

Restricting a sample based on known participant characteristics then using a random procedure to assign participants to each group.

Error variance or error

Variance attributed to or caused by the individual differences of participants in each group

Test statistic

Mathematical formula that allows researchers to determine the extent to which differences observes between groups can be attributed to the manipulation

Two independent sample t test

Test hypotheses concerning the difference in interval or ratio scale data between 2 groups

One way between subjects ANOVA

used to test hypotheses for one factor with two or more levels

Post hoc test

Computed following a significant ANOVA to determine which pairs of group means significantly different

Self reporting measure

items used in a survey, information is given to you BY the subject. It’s easy and cost effective but self report items are often inaccurate.

Behavioral measure

Speed and distance traveled by an athlete, behavior is recorded. Its more direct than self reporting but can require lots of ethical problems and some aspects of behavior are constructs.

Physiological measure

Physical responses of the brain and body like heart rate or body temp. When careful collection procedures are used these measures are unbiased but expense and training to operate equipment needed is expensive.

Null Hypothesis significance testing

inferential stats include a diverse set of tests of statistical significance called the NHST.

Null hypothesis

A statement about a population parameter, such as the population mean, that is assumed to be true BUT contradicts the research hypothesis.

AKA we begin by assuming we’re wrong.

Criterion

Probability value for the likelihood of obtaining data in a sample if the null were true for the population.

Retain (accept) the null

Greater than .05. Null results alone are rarely published.

Reject the null

Less than .05. Power is the probability that we will detect an effect if an effect actually exists in a population. this is the most publishable outcome.

P value

Probability of obtaining a sample outcome if the value states in the null hypothesis were true. Interpreted as an error.

Type II error

Probability of retaining a null hypothesis that is actually false. This means the researcher is reporting no effect in the population when there is one.

Type I error

Probability of rejecting a null hypothesis that is actually true. Researchers directly control for the probability of committing this error by stating the level of significance.

Goodness of fit test

Statistical procedure used to determine whether observed frequencies at each level of one categorical variable.

test for independence

Statistical procedure used to determine whether frequencies observed at the combination of levels of two categorical variables are similar or different from frequencies expected.

Effect

A mean difference or discrepancy between what was observed in a sample and what was expected to be observed in the population.

Estimation

A sample statistic is used to estimate the value of an unknown population parameter.