Illustrated Example of Exploratory Factor Analysis

Work Addiction Risk Test

The WART is a 25-item self-report inventory. Respondents are instructed to rate each item according to how well the item describes their work habits. Responses are scored on a 4-point scale: 1-never true; 2-sometimes true; 3-often true; and 4-always true. Summing the item responses across all items creates a total score, ranging from 25 to 100. A listing of the items on the WART is contained in Table 1.

Table 1

The Work Addiction Risk Test Items

Item

 

1

I prefer to do most things myself rather than ask for help.

2

I get impatient when I have to wait for someone else or when something takes too long.

3

I seem to be in a hurry and racing against the clock.

4

I get irritated when I am interrupted while I am in the middle of something.

5

I stay busy and keep many irons in the fire.

6

I find myself doing two or three things at one time such as eating lunch and writing memo, while talking on the phone.

7

I overly commit myself by biting off more than I can chew.

8

I feel guilty when I am not working on something.

9

It is important that I see the concrete results of what I do.

10

I am more interested in the final results of my work than in the process.

11

Things do not seem to move fast enough or get done fast enough for me.

12

I lose my temper when things don't go my way or work out to suit me.

13

I ask the same question over again, without realizing it, after I've already been given the answer once.

14

I spend a lot of time mentally planning and thinking about future events while tuning out the here and now.

15

I find myself continuing to work after my coworkers have called it quits.

16

I get angry when people don't meet my standards of perfection.

17

I get upset when I am in situations where I cannot be in control.

18

I put myself under pressure with self-imposed deadlines when I work.

19

It is hard for me to relax when I'm not working.

20

I spend more time working than on socializing with friends, on hobbies, or on leisure activities.

21

I dive into projects to get a head start before all phases have been finalized.

22

I get upset with myself for making even the smallest mistake.

23

I put more thought, time, and energy into my work than I do into my relationships with friends and loved ones.

24

I forget, ignore, or minimize birthdays, reunions, anniversaries, or holidays.

25

I make important decisions before I have all the facts and have a chance to think them through thoroughly.

Participants

Two different samples from previous studies (i.e., Robinson, 1999; Robinson & Post, 1997) were used for this study.  One of the samples (Robinson & Post, 1997) recruited participants from Workaholics Anonymous, a 12-step support group for work addiction, and conference registrants from national self-help conferences who had signed their names and addresses on a mailing list for participation in this study. Of 140 potential volunteers, 107 returned their questionnaires, resulting in a response rate of 76%. Two additional respondents did not complete all items on the questionnaire and were excluded from this study. The 105 respondents represented five regions of the United States and Canada. The mean age of respondents was 44 years with a range of 28 to 65 years. Sixty percent of the respondents were women. In this study this sample will be referred to as the workaholic group.

The second sample (Robinson, 1999) of participants was a convenience sample of students attending graduate and undergraduate classes at a large university in the southeast United States. A sample of 371 between 17 and 53 years of age (mean age 22 years) participated. A total of 2% (or 8 respondents) did not complete all the forms, yielding 363 in the final sample, 29% of whom were male (n=107), and 71% of whom were female (n=256). Participants were asked to complete the questionnaire during class time. In this study this sample will be referred to as the control group.


Running SPSS

[Download Data Here]

Move all 25 items to the variables list, then go to the Rotation option and select the Direct Oblimin option.

SPSS Output

Factor Analysis

Factor Analysis

 


 Results

    A exploratory factor analysis was used to evaluate the structure of the 25 items on the WART. A principal components method was used as the extraction method.  The decrease in eigenvalues (i.e., scree test) leveled off at five factors, each with an eigenvalue greater than 1.0. An oblique rotation (direct oblimin with delta equal to zero) was performed and retained for interpretation. The pattern and structure coefficients and communalities for this analysis are presented in Table 1.

Table 1

The Pattern Coefficients, Structure Coefficients, and Communality for the Items on the WART

Item

 

 

Factors
 
Communality

 

 

I

II

III

IV

V

 

1

 

-07  (15)

11  (20)

09  (17)

-74 (-74)

14  (16)

59

2

 

-06  (14)

65  (63)

02  (20)

13  (11)

03  (23)

42

3

 

54  (63)

36  (50)

04  (30)

17  (16)

-06  (16)

54

4

 

-04  (17)

67  (64)

-07  (14)

-01  (21)

02  (21)

41

5

 

75  (70)

-15  (09)

-03  (16)

-06  (08)

<01 (08)

52

6

 

74  (72)

04  (28)

-02  (21)

18  (04)

03  (18)

56

7

 

67  (69)

<01  (27)

16  (36)

24  (11)

06  (21)

55

8

 

62  (66)

02  (25)

-12  (19)

-20 (-30)

16  (26)

50

9

 

16  (25)

-02  (21)

-24  (03)

-15 (-14)

80  (78)

68

10

 

-14  (07)

02  (29)

23  (33)

-02  (02)

76  (79)

67

11

 

24  (43)

43  (61)

11  (35)

12  (06)

20  (40)

48

12

 

-20  (07)

61  (64)

18  (33)

06  (05)

14  (33)

48

13

 

07  (27)

12  (35)

47  (56)

14  (10)

21  (35)

41

14

 

21  (36)

20  (41)

24  (41)

21  (16)

25  (40)

39

15

 

49  (57)

02  (25)

20  (35)

-09 (-18)

<01  (13)

37

16

 

09  (36)

66  (70)

05  (29)

-23 (-28)

-06  (17)

56

17

 

14  (37)

63  (65)

-02  (22)

-21 (-27)

-10  (27)

50

18

 

50  (62)

38  (52)

-03  (25)

-01 (-12)

<01  (20)

50

19

 

61  (73)

07  (35)

14  (38)

-23 (-35)

01  (21)

62

20

 

53  (66)

13  (36)

23  (42)

-22 (-34)

-10  (07)

56

21

 

24  (40)

05  (31)

45  (57)

06  (01)

12  (26)

39

22

 

20  (41)

42  (55)

<01  (25)

-27 (-32)

12  (28)

44

23

 

32  (52)

03  (28)

46  (57)

-35 (-40)

-07  (07)

58

24

 

12  (32)

-13  (12)

75  (73)

-20 (-27)

-12 (-01)

62

25

 

-15  (12)

09  (29)

76  (75)

02  (01)

02  (16)

58

Note. Decimals have been omitted. Structure coefficients are in parentheses.


   
The total variance accounted for by the five factors was 52%. Pattern coefficients greater than .40 were used to determine relationships between the items and the factors. Using the major symptoms of work addition developed by Robinson and Post (1994) and a review of the factors by two psychologists with expertise in the area of workaholism, each factor was inspected for possible common themes among the items that loaded on the factor. The first factor had nine items with pattern coefficients greater than 0.40 (items 3, 5, 6, 7, 8, 15, 18, 19, and 20). These items appear to represent the major symptom of Compulsive Tendencies.  The second factor had seven items with coefficients greater than 0.40 (items 2, 4, 11, 12, 16, 17, and 22). All the items in this factor appeared to represent a major symptom of Control. The third factor consisted of five items (items 13, 21, 23, 24, and 25) and corresponded to the symptoms of Impaired Communication/Self-Absorption. The fourth factor had only one item (item 1) which appeared to represent Inability to Delegate. The final factor had two items (items 9 and 10) and seemed to represent Self-Worth. Item 14 did not have a coefficient greater than 0.40 on any factor. The factors and corresponding items are presented in Table 2.

Table 2

The Five Factors and Corresponding Items from the WART

Factors

 

 

Items

1. Compulsive Tendencies

 

 

3, 5, 6, 7, 8, 15, 18, 19, and 20

2. Control

 

 

2, 4, 11, 12, 16, 17, and 22

3. Impaired Communication/Self-Absorption

 

 

13, 21, 23, 24, and 25

4. Inability to Delegate

 

 

1

5. Self-Worth

 

 

9 and 10

Discussion

    The results indicate that workaholism, as measured by the WART, is not a unidimensional construct. The factor analytic findings suggest that the WART has five underlying dimensions: (a) compulsive tendency, (b) control, (c) impaired communication/self-absorption, (d) inability to delegate, and (e) self-worth. This is just one piece of information about the validity of the scores from the WART, and much more evidence is needed to demonstrate the viability of using the WART.