Lab Report 1 – The State-Trait Anxiety Inventory (STAI)

 

Psychological Assessment 566604    

Lab Programme 2016

Lab Report 1 – The State-Trait Anxiety Inventory (STAI)

The aim of this practical exercise is to introduce you to one widely used psychological assessment tool and also to introduce you to one of the most widely used statistical software packages in the world – SPSS.

Learning Objectives

After completion of this assignment you will:

Ø  Be able to describe the STAI including its purpose, administration, scoring and some important psychometric properties.

Ø  Be able to do basic statistical analyses of questionnaire data using SPSS including – descriptive statistics (mean, median, range, standard deviation, frequency distributions), independent samples t-test, scale reliability (Cronbach’s alpha and item-total correlations) and correlations.

INSTRUCTIONS:

• Your lab report must have a title page that includes: Name, Student ID, Paper title and Code Number, Title and number of lab report, date report is due, lab stream you are in, tutor’s name. Your name is optional if you prefer anonymous marking.
• Use Times New Roman or Arial font and use font size 12.
• Marks will be taken off for untidy presentation, poor grammar and spelling errors.
• There is no specific word limit as such however your lab report (not including the title page) should be no more than two (2) sides of A4 paper.
• To ensure that every student gets their unique results by conducting statistical analysis individually, you have to create your own, unique data set and use it for all statistical analysis in this lab report, as following:
• 1) Download the STAI Dataset from the assessment folder in Psychological Assessment
• 2) Complete the STAI questionnaire found in the same Assessment folder one more time.
• 3) Add your responses to each STAI item at the bottom of the STAI dataset.
• 4) Email your unique dataset with added responses to your tutor
• Successful completion of this task is worth 1 mark out of 25 allocated for this lab report.
• Failure to create and email your own unique dataset before the assignment due date means that your lab report will not be marked.

Answer each of the following 9 questions (total/25):

Q.1 Briefly describe, in your own words, the general purpose of the STAI and its administration (4 marks).

Q.2 Explain in your own words how the STAI is scored (2 marks).

Q.3 Using the STAI_PA_566604 dataset calculate the following (4 marks):

Ø  Mean, median, range and standard deviation for the total score of both State and Trait subscales of the STAI.

Q.4 A student score 35 on the Trait and 55 on State subscale of the STAI. Interpret these scores with reference to your answer to Q3 above. What does the difference between State and Trait scores mean? (3 marks).

Q.5 Examine the frequency distributions of the individual item response scores (1 – 4) for each of the 20 Trait items of the STAI (2 marks).

Ø  Describe the shape of the distribution typical of most of the 20 items and present onehistogram for one item to illustrate this.

Ø  You can cut and paste the histogram from the SPSS output.

Q.6 Conduct an independent samples t-test to compare total score trait anxiety of males and females and report the results appropriately including both means, t –statistic, degrees of freedom (df) and the p – value (2 marks).

Ø  Is there a significant difference between males and females?

Q.7 Calculate the individual item-total correlations for each of the 20 Trait items of the STAI (2 marks).

Ø  Report three items that have the highest item-total correlations including their values.

Ø  Report three items that have the lowest item-total correlations including their values.

Q.8 Calculate Cronbach’s α for both State and Trait subscales of the STAI, interpret and compare the coefficients (2 marks).

Q.9 Calculate both Pearson and Spearman bivariate correlations between State and Traitsubscales of the STAI, compare and interpret the coefficients (3 marks).

Ø  What coefficient is more appropriate to use with these data by considering the distribution of the total scores?

Ø  What correlation would you expect between state and trait measures and why?

 

 

eanwhile here are the more detailed explanations of the questions that need answering:

 

Lab Report : 

Q1.

STAII questionnaire: Number of items in each subscale, definition of state and trait, response options (scale of 1-4), reliability (refer to question 8/9), include reference (Julian, learning task blackboard), s

Q2.

Explain how scores range (2 subscales, 20 questions each) lowest score is 20, highest score is 80, what does that mean? Low score subscale means low anxiety, high = high anxiety, reverse score, what the items are reverse scores of items, list items, after scores were reverse coded, added to find final score??  1-20 state, 21-40 trait. Repeating is ok, need to answer whole question or you cant get the mark

Q3.

Check scores have been reversed (compare your data to blackboard), don’t include a table, two sentences. What is trait, what is state.  Be straight forward

Q4.

Compare q4 data to means, ranges, medians from question 3 The student scoring 35 on the trait subscale was approximately 1 standard deviation below the class mean. This means that the student has lower trait anxiety levels compared to the class average.

Difference between mean and student score

Provide example, why the student might have higher scores/ anxiety. Student had higher state anxiety than to class, this could be because the student was surprised with an exam which was why the score was higher

Q5,

Looking for distribution most typical of most items, find pattern.

How many were positively skewed, normally distributed and negatively skewed

Ten of my histograms were normal distribution (item numbers 1,5,7) Sixt were positively skewed (item numbers xyz) Four were negatively skewed (Item number xyz).

Provide a histogram that is typical of most of your items. Use captions

Ten of histograms were normally distributed because the mode was 2

Caption eg

Figure 1: Histogram showing distribution of student scores for item 1 on the state subscale of sTAII (n=73)

No colour: double click on graph; fill in boarder, select grey, black or white, change S1 to ordinal scores on x axis, frequency on y axix,

Q6.

Online paper- Learning tasks, independent samples t-test, model answer (No significant difference was found between men and woman in terms of their…..), follow format of showing t, report mean (SD=)

Is there  a significant difference between males and females

Turn table into words, report in lab report: mean total for males 40.82 (SD=10.71) and the mean for females is 45.73 (SD=9.69)

Levenes Test sig (significanc/ pvalue) must be above 0.05 (opposite to pvalue) to use a t-test. (use top line of the table if so)

T(70= -1.773, p= .08)

The independent sample found there was a significant difference found / no difference found

T(Degrees of freedom (df) = t score (t) p=(Sig. 2-tailed)

Q7.

Analyse, scale, reliability analysis, move across Trait items, Statistics, item scale and scale if item is deleted, alfa, ok

Item total statistics, Corrected item total correlation,

If there is a low value the item is not fitting well in the scale, might not be measuring anxiety,

Report three items that have highest and lowest values

Baresline reliability is .7

Item total Statistics (corrected item total correlation) report 3 items with highest correlations
Ignore negatives (still use the numbers but ignore the negative symbol)

Item (10 write question), .75

Item

The items that had the highest total correlation were item 10 (I remained curious about the nature of each experience as it arises ) .747. Item 5 (I was curious to see what my mind was up to from moment to moment) .75 and item 15 (I noticed various sensations caused by my surroundings (eg., heat, coolness, the winf on my face)) .73

The items that had the lowest total correlation were item 6 (I was curious about each of the thoughts and feelings that I was having) .29, item 13 (I was curious about what I might learn about myself by just taking notive of what my attention gets drawn to) .42 and item 14 (I noticed emotions come and go) .45

Q8.

Cronbach’s Alpha (Analyse, scale, reliability analysis, move across Trait items, Statistics, item scale and scale if item is deleted, alfa, ok)

Reliability of scale has to be .7 or higher, should not be used on general population

Cronbach’s Alpha for the sub stait scale of the STAII has interpreted 0.91. This is a strong value and this scale is reliable as the value is higher than 0.7

This is a very reilable scale (is the state or trait scale more reliable)

The lowest acceptable value is .70 thereforre both subscales have high reliability.

Compare: Trait subscale has a lower reliability coefficient than the state.

Q9

Ordinal level data = use spearman, but if data is normally distributed use pearson

Because the data is ordinal we should use spearman, however because the data is normally distributed we could technique use Pearson. Looking at both values the Pearson and Spearman coefficient are very similar therefore we could use either.

The correlation was .70 (p = .04). The significant correlation could be due to the state and trait measuring some of the same things, as they both measure anxiety. However they have some differences in what they measure as state and trait are different in some ways.

0.00 is significant

P= sig (2-tailed)

Seeing as distribution of scores are normally distributed I chose to you the pearson correlations

Correlation between the two groups was 0.00 meaning there is a difference between the two groups

The correlation reported between state and trait is 0.678 (how strong relationship) with a pvalueof(0.00). This shows that there is a significant difference between the two scales.

Non-normal distribution use non-parametric scale = spearman