Writing an Introduction / Personality & Risk Taking Lab 6kvampete/Lab 6.pdf · Writing an Introduction / Personality & Risk Taking ... data. –Calculate ... Writing an Introduction

Writing an Introduction / Personality & Risk Taking

Lab 6

PSY 395 Fall 2013

Descriptive Statistics and Histograms

• You should get in the habit of taking a peek at the data.

– Calculate relevant descriptive statistics and take a look at how a variable is distributed. Does it look normal?

You can get Descriptive statistics and a Plot with one function:

Descriptive Statistics Frequencies

A frequency window will pop up.

Drop the scale scores into Variable(s) Then Choose

Statistics

Check off the measures of central

tendency & dispersion you want and

click continue.

Then choose Charts

Select

Histograms &

Check mark

“With normal

curve”

You will get the descriptive stats in

Output.

Ask yourself: Is this the mean and

standard deviation I would expect?

You will get a

histogram and

normal plot. Do the

data look normally

distributed?

• We often want to know if two variables go up and down together or if one goes up while the other goes down.

– Is alcohol consumption (Y) negatively related to heart disease (X)?

– Does the accuracy of performance (Y) decrease as speed of response (X) increases?

– Are people who are high in extraversion also high in openness?

Correlation

Pearson’s r

• Measure of the size of a linear relationship between two variables.

• Ranges from a perfect negative correlation -1.00 to a perfect positive correlation +1.00

– If r = 0 then no linear relationship exists.

r =cov(x, y)

sxsy

Covariance: Degree to which x and y

vary together

Degree to which X and Y vary

separately

Calculate a correlation with SPSS

• Let’s calculate the correlation between extraversion (EXTRA) and openness (OPEN).

• Analyze Correlate Bivariate

Drop EXTRA and OPEN into the Variables window.

The output produces a correlation matrix

• The table also tells you if the correlation is significantly different from 0 and how many observations were used to calculate the correlation.

These are the

correlations between

Extra and Extra.

Notice they are equal

to 1.

These are the

correlations between

Extra and Open.

Notice they are the

same.

Tables

Note only horizontal lines are

used to organize the table.

Basically only a bottom and top

line around the column labels.

And a line at the bottom.

Make it easy for the reader to

compare the data you want

them to compare.

• Recall, reliability is a measure of how consistent a test score is. Different types of consistency:

•test-retest

•internal consistency (alpha)

•interobserver/interrater

Reliability

• Quantifies how much consistency there is between the items in a scale

–How well do the items “hang together”?

• If different items in a scale are consistently measuring the construct of interest, we should find associations between all of the items…

Internal Consistency

• The first step is to look at the Inter-item correlation Matrix

• For example, we can examine the associations between each of the items that comprise the Extraversion scale.

• Under Analyze

–Correlate Bivariate

–Select all items in the Time 1 Extra scale (11R, 23R, 2, 16, 17, 32, 41, 43, 45, 49)

–Click OK

Internal Consistency

Drop the extraversion

items into the Variables

box.

• Recall the Extraversion items are:

• bf11R, bf23R, bf2, bf16, bf17, bf32, bf41, bf43,

bf45, bf49

To assess internal consistency: Run correlation between all the scale items

Analyze Correlate Bivariate and this window will open

This is the correlation between

item 17 and the reverse code 11r 17: I enjoy being part of a loud crowd

11: I dislike loud music

If you want to look up what the content of each item is,

remember the personality items are on Angel!

16: I am willing to try anything once.

Remember the

correlations are the

same above and below

the diagonal!

These are the correlations

between items 16 and 17 and

the reverse code 11r

17: I enjoy being part of a loud crowd

11: I dislike loud music

16: I am willing to try anything once.

Remember

the

correlations

are the same

above and

below the

diagonal!

• Examine the matrix carefully. What is the range of correlations? Are there any particularly low values? Any particularly high values?

• Download the excel sheet with the Personality Items and take a look at the individual items. Do the high inter-item correlations make sense? How about the low ones?

Calculating Alpha by Hand

a =kri, j

1+ k -1( )ri, j

k: number of items average inter-item

correlation

ri , j :

Calculate Alpha

• Say the average inter-item correlation for our extraversion scale was = .27. What is Alpha for k = 10?

• Now imagine we had the same inter-item correlation but now calculate Alpha for

– k = 20, k = 40, k = 80, k = 160 ri , j

ri , j

Calculate Alpha

• If the average inter-item correlation for our extraversion scale were = .27. What is Alpha for k = 10? Answer = .79

• Now imagine we had the same inter-item correlation but now calculate Alpha for:

– k = 20, k = 40, k = 80, k = 160

– .89, .94, .97, .98 ri , j

ri , j

Why does Alpha approach 1 as you add more unique items?

• Remember our measures of reliability are telling us how much measurement error there is. As we add more unique items then we reduce the impact of measurement error.

• But, no one wants to take a 1000 item inventory!

Calculate Cronbach Alpha with SPSS

Drop the items for the scale of

interest (e.g., extraversion) into

the item box.

Choose Alpha

And type in a

Scale Label

Want to know a trick?

You can get a lot of good stuff out of the statistics option. Click it

and take a look.

You can get:

inter-item correlation matrix

mean inter-item correlation

means and sd for each item

Output for reliability looks like this.

The second table has Cronbach alpha,

which is .83.

• Examine the inter-item correlation matrices for the other Time 1 Big Five scales. – Openness (bf9, bf35, bf39, bf50, bf5R, bf18R, bf21R, bf24R, bf38R, bf44R)

– Conscientiousness (bf19, bf25, bf40, bf10R, bf14R, bf20R, bf27R, bf29R, bf36R, bf42R)

– Extraversion (just performed in the slides above)

– Agreeableness (bf6, bf8, bf12, bf28, bf31, bf46, bf3R, bf4R, bf13R, bf15R)

– Neuroticism (bf7, bf22, bf30, bf33, bf48, bf1R, bf26R, bf34R, bf37R, bf47R)

Internal Consistency

You will need this for your table in your Lab Report 3!

Construct Validity

• Does a particular measurement truly measure the construct?

– THINK: OPERATIONAL DEFINITION.

• The extent to which the variables accurately reflect or measure the behavior of interest.

Assessing Construct Validity

• Correlate the measure with measurements of similar but distinct traits (divergent validity).

– Correlate extraversion with agreeableness, extraversion with openness, etc., they should not be strongly correlated.

• Could also compare the measurement with other measures with the same trait.

– Correlation extraversion from the Big 5 inventory with extraversion Eysenck’s personality scale.

In SPSS correlate the 5 scale scores

Drop the scale scores

into the variables box.

There is a -.65

correlation

between

contentiousness

and extraversion.

Why do you think

there is a

significant

correlation?

Remember in a

correlation matrix, the

correlations are the

same above and below

the diagonal.

Like extraversion and

contentiousness here.

NOTE!!! This is an example of what your output should look like. We can’t give you everything!

Criterion Validity

• Can a measure accurately forecast some future behavior, or

• Is a measure related to some other measure of behavior?

• In our study, our criterion is risky decision making in the BART.

For our criterion, we will use the BART

• Recall, we are going to use the average number of pumps taken on non-exploding balloons (adjusted BART score) as a measure of risk taking.

Guess what? Run another correlation! Analyze Correlate Bivariate and this should pop up

Drop the scales

into the Variables

box. Also drop

“adjBART” in at

the bottom.

The bottom row will show the relationships between the scales and the

adjusted BART scores. Anything pop out at you?

NOTE!!! This is an example of what your output should look like. We can’t give you everything!

For Lab Report 3

• Do a lit search for three peer-reviewed articles to motivate your hypothesis about an association between a personality factor and risky decision making.

For your lab report

• Report in a correlation table the inter-correlations between the personality scales and the adjusted BART scores from the BART.

– List the correlations between the diagonal.

– List the reliability (Cronbach Alphas) of the scales along the diagonal.

Here is what the table should look like.

References and in text citations

• When you write a paper you need to cite the sources you are drawing your ideas from, using or adapting methods from, using or adapting analyses from, etc.

References & in text Citation

MAKING ASSESSMENTS WHILE TAKING REPEATED RISKS 3

Making Assessments While Taking Repeated Risks:

A Pattern of Multiple Response Pathways

Risks like texting while driving, eating unhealthy food for a quick lunch, or smoking a

cigarette, are typically not one-shot choices. Rather, these are choices that are made repeatedly

over the course of a day, a week, a month, and so on. Repeated decisions put different demands

on a cognitive system than one-shot choices. They, for example, allow for and even may require

learning (Busemeyer & Stout, 2002; Denrell, 2007; Pleskac, 2008; Wallsten, Pleskac, & Lejuez,

2005) as well as search and exploration (Daw, O'Doherty, Dayan, Seymour, & Dolan, 2006).

They also put different demands on attention and memory (Barron & Erev, 2003) and may even

change the decision process itself (Jessup, Bishara, & Busemeyer, 2008). In this paper, we

examine how multiple response pathways develop while people make repeated decisions.

Judgments and decisions under uncertainty are often described as being made with one of

two information-processing systems: System 1 or System 2 (Evans, 2008; Kahneman, 2003;

Mukherjee, 2010; Reyna, 2004; Sloman, 1996; Stanovich & West, 2000) (cf. Gigerenzer &

Regier, 1996; Keren & Schul, 2009; Kruglanski & Gigerenzer, 2011). System 1 makes

judgments that are fast, automatic, effortless, associative in nature, and undemanding on

computational capacity. System 2, in comparison, makes judgments that are slower, controlled,

rule-based, and demanding of computational capacity. Cognitive theory tells us that the

development of an automatic response process–one characteristic of System 1–often requires an

appreciable amount of repeated and similar actions (Schneider & Shiffrin, 1977; Shiffrin &

Schneider, 1977). Thus, making repeated decisions should be sufficient for the development of

an automatic response.

In text

citation

s look

like

this: Referen

ces look

like this:

References

• https://owl.english.purdue.edu/owl/resource/560/05/

• http://libguides.lib.msu.edu/citeinfo

• The format changes depending if you have: – A journal article with one author

– A journal article with more than one author

– A magazine

– A book

– Electronic sources

General Structure of References

Authors (Year). Article title.

Journal Title, Volume, Pages.

doi:

Hanging Indentation – After the

first line of each reference entry

subsequent lines should be

indented one-half inch from the

left margin (Purdue Owl, 2013).

DOI’s • Digital Object Identifiers: unique alphanumeric identifiers that lead

users to digital source material

– They are like a paper’s email address

Here is what a reference section looks like in an unpublished report

Lab Report 3

• Due by the start of Lab during the week of October 28th.

• READ ALL DIRECTIONS!

• Ask me – your amazing and friendly TA (Weaver, 2013) – if you have any questions.