PSY 395 - Lab #1, 1

PSY 395 - Lab #1

First thing…let’s collect some data!

This is a measure of normal personality – not abnormal personality (e.g., the MMPI), so there are no incorrect answers. Please provide responses to these items (#1-112) using the provided scantron.

Before you complete the form please do the following now:

So that you get homework credit for doing this, please put your PIDthe self and peer scantrons now so you don’t forget.

The IPIP (International Personality Item Pool) is a self-report questionnaire that measures 5 personality dimensions (the Big 5). You can learn more about the IPIP at

  • Extraversion: Sociability, assertiveness, talkativeness, activity. Opposite is introversion.
  • Agreeableness: courteous, flexible, trustworthy, good-natured, cooperative, tolerant
  • Conscientiousness: dependable, careful, thorough, responsible, hard-working, perseverance
  • Neuroticism: anxious, depressed, angry, emotional swings, insecure, embarrassed. Opposite is termed emotional stability.
  • Openness to experience: imagination, cultured, curious, original, artistic sensitivity

The IPIP test consists of 100items, with 20 items for each of the Big Five traits (e.g., 20 items for Extraversion, 20 items for Agreeableness, etc.).

Typically for psychological tests that are self report:

  • A person is asked to give a response for each item (e.g. "yes" or "no", "strongly agree", "never", "female").
  • Then, a numerical value is assigned to that response, for example:
  • "yes" = 1, “no” = 0;
  • "strongly agree" = 5, “agree”=4, “neutral”=3, “disagree”=1, “strongly disagree” = 0
  • “male” = 1, "female" = 2
  • In other words, there is a numerical value - a score - that is assigned to each person’s response to each individual item.

Items are then used to make scales

The scores of individual items, which are thought to measure the same thing (construct, such as Extraversion), are added together to make a scale that measures that construct.


For example…On the IPIP, there are fivescales.

Each scale consists of 20 individual items

The scores for the 20 items are combined (added) to give us a scale score – one for each of the five scales.

Seems simple, right? Item scores, add together, get a scale score.

-Just one teeny complication…reverse scoring of some of the items.

For many scales in psychology, the higher the person is on the construct that the scale is supposed to measure, the higher the person’s score is on the scale. This is true on the IPIP for the 5 personality scales.

For example, if Timmy, got a score of 15 on the Neuroticism scale and Jimmy got a score of 30, Jimmy would be considered to have a higher level of neuroticism (be more neurotic) than Timmy.

But first let’s look at the items on the IPIP - not all the items are worded in the same ‘direction’.

For the positively worded items, the highest score on the item corresponds to the highest amount of the construct the item is measuring.

Example: Item 1 on the IPIP - "I am the life of the party" - is an Extroversion item.

A higher score on this item = higher level of extroversion.

But there are also negatively worded items, where the higher a person’s score on the item, the lower the person is on the construct the item is measuring.

Example: Item 6 on the IPIP - "Don’t talk a lot" -is also an Extraversion item.

A higher score on this item = lower level of extraversion (because higher extraversion means more sociability).

So before a scale is created (before you add the scores on all items together), the scoring on all the items have to go in the same direction, where a higher score on each item = higher on the construct being measured.

You do this by reverse scoring - changing the scores on the negatively-keyed items to their opposites

For example, if you are an Extravert your scores across the items should be high (more toward 5 than toward 1). But take Extraversion item #6, “Don’t talk a lot.” An Extravert would answer with a '1' (‘Very Inaccurate’). This clearly is not in the same direction as the other items such as “Likes parties” where an Extravert would answer ‘5.’

…so, this score of '1' would be changed (reverse scored) to a score of '5' before the all the Extraversion item scores were added to make the scale (and any scores of '4' would get reverse scored to a '2').

Okay, so we have reversed the items that need to be, and now we have scales,so what kind of scale is it?

There are four types of measurement scales for data in psychology…

  • Nominal or categorical scales (unordered categories) - Numbers serve as labels for different categories
  • e.g., gender (male/female), hair color (blond/brown/black/red/grey/bald), mental health categories (depressed/bipolar/other)
  • Ordinal scales (ranks) - Scores represent ordering or ranks. Here there is no information about differences or distances between scores (e.g., you know someone was the 2nd child but you don’t know from this information how much older the 1st child is)
  • e.g., birth order, order of finishing in a marathon
  • Interval scales (continuum) - Scores represent real amounts that reflect relative differences in magnitude.
  • e.g. SAT scores, temperature
  • Ratio scales (continuum with an absolute zero) - Like interval scales, but they have a true zero point. Differences between scores have absolute and relative meaning. For example, if Mary scores twice as much as Sherry, Mary has twice as much of the quality being measured than Sherry has.
  • e.g. pounds, height, reaction time

Note: Nominal and Ordinal scales cannot be used to find the mean, standard deviation, or variance. (e.g., After all, what’s the mean of 3 red-haired and 2 black-haired adults?). You could take the mode for nominal scales, though (e.g., you find that most individuals have red hair). Most standard statistical procedures in psychology require an interval or ratio scale.

Scales of Measurement: Practice

What kind of data are these (nominal, interval, ordinal, ratio)?

Weight (in pounds)

Favorite Color: 1=red, 2=blue, 3=yellow

Class Rank

Age

SAT score

IPIP scales

Can you come up with one more example each of nominal, interval, ordinal, and ratio?

One type of interval scale, called a Likert scale, is used a lot in psychological measurement (when measuring attitudes, personality, etc). A Likert scale is something like using a ruler to for measuring a psychological construct. Remember that constructs are abstract – we invent constructs based on theories of what, psychologically, resides within a person. For instance, you can’t see someone’s level of conscientiousness directly - you have to measure itindirectly.

  • Responses generally fall along something like a 1 - 5, 1 - 7, 1 - 9, etc. Likert rating scale
  • Two types of response scales:
  • Graphic scales - just the end points are defined

Strongly Disagree Strongly Agree

1234567

  • Numerical scales - each response choice is defined

Strongly Moderately Somewhat Neither Somewhat Moderately Strongly

DisagreeDisagree Disagree Agree nor Agree Agree Agree

Disagree

1 2 3 4 5 6 7

Never true Rarely true Sometimes true Often true Very often true

12345

Some basic points about rating scales:

  • Typically the reliability of a rating scale increases as the number of response options increases. For example, a 1 - 9 scale is typically more reliable than a 1 - 5 scale and certainly more reliable than a dichotomous scale which where you respond 0 or 1 (e.g., yes/no, or like me/not like me).
  • An odd number of response choices allows for a neutral (midpoint) category; e.g., on a 1-5 scale the neutral point is a 3, but on a 1-4 scale there is no neutral point and a test-taker must take a ‘side’ one way or the other.
  • Problem: There might be biases due to response styles or response sets – Examples: some people choose the extremes, while others avoid them; some people are more agreeable (or disagreeable) regardless of the item; sometimes the order of the items influences how people respond

Assigning numbers to peoples’ item responses allows us to judge

(a) the consistency or accuracy of the items and scales – this is reliability – in other words, how close are individuals’ observed scores to their true scores (remember Classical Test Theory)?And

(b) the relationships between the scales with different outcomes of importance – this is validity– do the scales make predictions that support a research hypothesis (e.g., Conscientiousness predicting GPA).

Generally, computers and statistical programs (vs. a chalkboard and abacus) are used to do answer questions of reliability and validity.

  • In this class, we will use SPSS (Statistics Package for the Social Sciences)
  • We’ll use the term variable a lot of the time to refer to the items or the scales that we enter into SPSS
  • A data set consists of a file where a case (person who took the test) is in each row and a variable (item score or scale score) is in each column.

Intro to SPSS Statistical Analysis
(also refer to the SPSS Help Menu – it has a lot of information)

  • How to do stuff in SPSS
  • Opening files:
  1. Open the SPSS program
    (your TA will help you find it)
  1. A menu might come up – click “Cancel” and you will see a blank data set.
  1. On the File menu you will see ‘Open’ so that you can open SPSS files you are given. If you go to ‘Open’ then you will be asked to go to where the file you’d like to open is (e.g. your floppy disk, your Public AFS space, etc.)

Once you either open an SPSS data file, or you type information into an SPSS data set, you can go to ‘Save’ under the File menu to save your data to disk:

  1. SAVE OFTEN AS YOU’RE WORKING, AND ALWAYS BEFORE YOU CLOSE THE PROGRAM.
  1. Go to the File menu; choose ‘save’
  1. Choose where you’d like to save your file (e.g., your floppy disk, your AFS public drive). Make sure you save in the right directory.
  1. ALWAYS SAVE YOUR DATA FILE. SAVE TWO (2) COPIES OF YOUR DATA (AND YOUR HOMEWORK AND ANY OTHER COMPUTER WORK) IN CASE ONE GETS LOST OR CORRUPTED (SEE SYLLABUS).

Notice there’s a lot of bold print here – because saving and backing up your work is very very important! 

  • Creating variables and entering data directly into an SPSS data file:
  1. Open SPSS and click cancel if you get a window prompting you to open a file.
  1. You will now see a data sheet with a series of empty columns and rows.
  1. On the first row, list the names of your variables.
  1. For example, the first variable might be ID (a code number designating each participant).
  2. Look at the bottom of the window and you will see a tab called “Variable View.” Click on this tab to enter the name of a variable. In Row 1 under “Name” you could type “PID” and that would be the name of the first variable (don’t type the quotation marks, just type what is in quotes). Then in Row 2 under “Name” you could type “e1” which is an item measuring Extraversion, and it is item #1 on the IPIP. Notice it automatically tells you the data are “Numeric” – you could change this but there is no need (your data are numeric).
  1. After naming your variables, click on “Data View” and proceed to enter data into the data sheet, one row for each subject. For practice you will only enter your own data.

SPSS Practice with Your Own Data

For Homework 1…

Go to Variable View and create your Variable Names

  • First variable is "PID" –the last five digits of your PID number
  • Second variable is the first item on the IPIP, which is a Extraversion item, so type "e1" into the second row.
  • Third variable is the second item on the IPIP, which is an Agreeableness item, so let's call it "a2"
  • Fourth variable is the third IPIP item, which is a Conscientiousness item, so we'll call it "c3"
  • Fifth variable is an Neuroticism item, "n4"
  • Sixth variable is an Openness item, "o5"
  • Seventh variable is another Extraversion item, so let’s call it “e6”
  • Eight variable is another Agreeableness item, so let’s call it “a7”
  • …and so on through “o50” (stop at the 50th item)

The pattern should be somewhat clear - the items go in order of the scales – E, A, C, N, and O – so every fifth item belongs to the same scale (e.g. the sixth IPIP item is another Extraversion item, the seventh IPIP item is another Agreeableness item, etc.)

EnterONLY THE FIRST 50 IPIP items for your test.

After the first 50 items of your own data are entered, you will create your score for the Conscientiousness scale. But before you do that you'll need to reverse score some of the conscientious items. So…

Reverse scoring in SPSS

  1. Recall that reverse scoring negative items is "reversing" the item's scores (e.g. 1s become 5s, 2s become 4s, and 3s stay the same)
  2. For example e6 (second Extraversion item on the test) is a negatively-worded item. To create a reverse-score variable for it, go to the menus at the top and click on “Transform” and then “Compute”.
  3. There is a blank for “Target Variable” – this is the name of your reverse-scored item. Let’s call it “e6_re” (6th item, Extraversion scale, reverse scored) so enter that in the blank.
  4. Now after the equals sign, type “6 – e6” (don’t type the quotes). What you are doing is taking six minus your score for e6. Do you see how this is reverse scoring? If you said “5” to e6, then what will e6_re be? If you said “1” to e6, then what will e6_re be? What is you said 3?
  5. You are only reverse-scoring your scores, which is one row of scores, but if you were working with a larger data set, what you just did would reverse-score the item for everyone in your data set (all N rows of scores for the N people in class).
  6. This was an example for one Extraversion item; you will now do this for all the Conscientiousness items that need to be reversed. (e.g., item c8, c18, c28, c38 and c48). Remember you are only working with the first 50 items of the IPIP that you entered. So, there are only 10 Conscientiousness items, and 5 of those need to be reversed. When you get the data file for everyone, however, that will be based on all 100 IPIP items, where there are 20 Conscientiousness items and 10 are reversed to create the Conscientiousness scale score.

After all the items are entered and reverse scored, you can compute a scale in SPSS

  1. Use the Transform/Compute menu options that you used to reverse score the Conscientiousness items. Name the Target Variable “consc” to be the Conscientiousness scale score.
  2. Now you are going to add up the items that make up the Conscientiousness scale –except, for items that were reverse scored, make sure you use the new variables labeled with “_re” suffix and don’t use the original variable (remember, the original variables was in a direction opposite to the other items on the scale – which is why you reverse-coded in the first place!).
  3. So you’ve labeled the “Target Variable” as “consc.” Now for “Numeric expression” you want to have “c3 + c8_re + c13 + c18_re + c23…” and so on out to c48. You can just type this in the window on the right, or you can do this all by clicking: click the relevant item label in the bottom left column, then click the arrow that points to the right so the name jumps into the “Numeric Expression window”, then click the plus sign and a plus should show up in the window, and keep repeating this process until all the Conscientiousness items have been added together.
  4. Press “OK” and the scale score will be computed. The variable shows up as the last column of the file. Save this file.

Homework 1

due at the BEGINNING of your next lab

(no late assignments accepted)

  1. Fill out your scantrons (done in class)
  2. Create variable names in SPSS for only the first 50IPIP items
  3. Enter your scores for only the first 50IPIP items
  4. Reverse score the appropriate items from the Conscientiousness scale
  5. Compute the Conscientiousness scaleon items contained in the first 50 IPIP items

(items 3,8,13,18,23,28,33,38,43,48)

  1. Send final SPSS data file (variables entered, with reverse scored items, scale) to your TA as an e-mail attachment.
  2. You will be given an IPIP peer measure/scantron to a friend to fill out. Be sure to put YOUR PIDon it so that you receive credit. Turn this in at the beginning of your lab session next week (don’t forget to have your peer fill this out and return to you before next week - this is an important part of your homework grade). Late scantrons will not be accepted.