Test 3

Structure of Lit review

General to specific

Topical order

Variable A → Variable B → The why

Synthesis of information

No string of research, make a claim instead

Plagiarism

DO NOT use the exact words, use quotes and a citation

Academic voice

Content, not entertainment

Goal of qualitative data analysis

To introduce the basic principles of statistics and the basic statistical tests so that you can be a competent consumer

Data analysis

The process of examining what data means to the researcher

Goal of data analysis

Transforming data into useful information that can be shared with others

Key of data analysis

Reduction

Descriptive statistics

Used to summarize the info in a given data set pertaining to a particular sample

Inferential statistics

Going beyond the data, making inferences and drawing conclusions with greater or lesser degrees of certainty making generalizations back to the populations

Dispersion

How spread out our scores are and how much variation we have within those scores

Frequencies

Where the data lies

Central tendencies

The center points

Mode

Most frequent score (Most appropriate for nominal data)

Can have multiple or none

Median

The middle most score in the distribution (Appropriate for ordinal data)

Mean

The average of all scores (Appropriate to interval/ration data)

Range

Reports distance by reporting the highest and lowest scores

Standard Deviation

Average deviations of the mean expressed in the original unit of measurement

Helps us understand the distribution of scores

Normal distribution

Bell curve

Represents the location of deviations about the mean and the probability of these different deviations occurring

Standard scores

A common unit of measurement that indicates how far any particular score is away from the mean

They locate scores within a distribution

Z score

Z= x-m/SD

Frequencies

Counting the occurrence of a score or item

Nominal frequencies

Frequency

Frequency percentage

Interval/ratio frequency

Cumulative frequency

Cumulative frequency percentage

Visual displays of frequencies

Pie charts

Bar charts

Histograms

Polygons

Density curves

Histograms

Bar chart thats connected

Polygons

Uses dots and lines that connect these dots

Density curbe

For large amounts of data, bell curve

Inferring from data

Taking your result and determining the likelihood of it happening by chance

Inferential statistics

Goes beyond descriptions to draw conclusions

2 main purposes of inferential statistics

Estimation

Significance testing

Estimation

Estimating population characteristics

Significance testing

Hypothesis testing

Using statistics to draw conclusions to see differences in groups

Key to inferential statistics

Randomness and probability

Parametric statistics

Based on normal distribution

Good estimates using parametric statistics

Normality assumption

Random sample

Normality assumption

The variable of interest is normally distributed in the population

Goal of estimations

To estimate populations using parametric statistics

Central limit theorem

How do I get a sample that is representative of the population

Abnormal distributions

Kurtosis

Skewness

Kutosis

Pointed or flat

Skewness

Assymetrical

Leptokurtic

Pointed and tall distribution

Platykurtic

Flat distribution

Bimodel distribution

Multiple tops

Positively skewed distribution

Tail is on the right

Negatively skewed distribution

Tail is on the left

Hypothesis testing

Examining how likely differences between groups or relationships between variables occur by chance

Larger difference/stronger relationship

Low likelihood of chance

Level of significance

Represents the probability that our results are due to sampling error (chance)

Common significance levels

.05, .01, .001

Expressed level of significance

P < .05

Lower P-Value

More sure that there is a difference not by chance

Key to significance

Tells us if there is a result we can interpret

Null hypothesis

A statement that statistical differences or relationships have occurred for no reason other than chance

Possible errors with null hypthesis

Type I error

Type II error

Type I error

Rejecting a null hypothesis that in reality is true

Finding something that is not there

Type II error

Accepting the null hypothesis that in reality is false

Not finding something that is there

Procedures for hypothesis testing

Set significance level

Select statistical test

Compute statistic/value

Compare value to critical value

Region of acceptance; region of rejection

Accept or reject the null

Power

The probability of rejecting a null hypothesis that is in fact false

The likelihood of finding something that is there

Increasing power

Increasing sample size

Procedure for all hypothesis testing

Set significance level

Select statistical test

Compute statistic value

Compare to critical value

Region of acceptance/rejection

Accept or reject null

 χ2

Used with nominal dependent variable

Not based on normal distribution

Types of  χ2 tests

 χ2 for goodness of fit

 χ2 for independence

 χ2 for goodness of fit test

Determines how well the frequency distribution for a sample fits the population distribution

 χ2 for indepence test

Tests whether or not there is a relationship between two variables

Rejecting the null/signigicant

Stat(t) > CV

Accepting the null/non significant

Stat(t) < CV

Critical Value

Value needed to achieve statistical significance

t-test

To see if two means significantly differ

Types of t-tests

One sample (compared to population)

Independent

Paired

ANOVA

F-Test

Same thing as a t-test but can have more than two groups

Association

Not causal, implies there is a relationship

Advantages of of ANOVA over t-test

Compares 2 or more groups

Controls for when error adds up

Can have multiple IV

r

Independent observations

Helps us understand linear relationships

r provides two pieces of information about linear relationships

Direction of relationship

Magnitude of relationship

Direction of relationship

Positive number, positive relationship

As one variable goes up, the other does as well

Magnitude of relationship

Square r

Slight magnitude

Low magnitude

Moderate magnitude

.36-.6

High magnitude

.61-.8

Very high magnitude

>.8

Correlation coefficient

Interpreting r

* not causation
* restricted range
* outliers

Effect sizes

Assess the degree of interdependence between two variables

Always expressed as a number of 0 and 1

Correlation

R^2

ANOVA

η²

Statistic for effect size

η²

Significance and effect size

We want both to be big

Emergent

We learn as we go, it emerges as the info is gathered

First order explanation

The participants attitudes, beliefs, etc

Second order explanation

The researcher imposes their understanding of the topic

Confirmace

Conclusions due to the data and not the researchers bias

Credibility

Conclusions ring true with the participants

Process of qualitative data

Questions to frame/guide analysis

Unitizing

Developing coding strategies

Plug holes

Checking for conformability and credibility

Finding examples

Integrating categories

Unitizing

Break up into smallest pieces of information that can stand by itself (meaning nuggets)

Developing coding categories

Constant comparative method

Constant comparative method

Read all the data (Immersion)

Give each unit a code and categorize it by comparing it to other units

Create category labels and descriptions

Repeat coding process

Plug holes

Gather more data until saturation