Data Analysis

UNIVARIATE ANALYSIS

Frequencies and Reliability Dependent Variables

Scale: SOCIAL DISTANCE

· Six items asking about willingness to interact with the mentally ill ranging from having a treatment center in the neighborhood to willing to have children with a person who has mental illness. Answers scored yes and no. All yes scores were added together. The highest possible score is 6 suggesting a willingness to interact; the lowest possible is 0. The higher the score the more willing to interact..

· Based on Bogardus Social Distance Scale

· Average score is , 4.09 suggests average person in sample willing to interact

· Reasonable reliability Cronbach’s alpha .690.

Scale: ATTITUDE TO THE MENTALLY ILL

· Nine items asking about attitudes to the mentally ill, extent to which negative stereotypes held including dangerousness, ability to keep job, intelligence. Scored 1-9, with higher score indicating more agreement with these items. Higher score indicates more negatively held stereotypes. Highest possible score 91; lowest 9.

· Based on scale created by Corrigan

· Average score is , 25.57 suggests average person in sample holds relatively few stereotypes

· Reasonable reliability Cronbach’salpha ..853

Scale: OPENNESS TO TREATMENT

· Six items asking about openness to treatment for mental illness. Includes items such as “There are certain problems which should not be discussed outside the home.” Scored 1-5 with higher score indicating more agreement with these items. Possible scores 6 to 30. Higher score indicates less openness to treatment.

· Based on scale created by Corrigan

· Average score is 11.63 suggests average person in sample relatively open to treatment

· Reasonable reliability Cronbach’s alpha .739

Scale STIGMA

· Five items asking about the extent to which the respondent feels shame at personally expressing mental illness or going to treatment. Scored 1-5 with higher score indicating more agreement with these items. Possible scores 5 to 25. Higher score indicates greater stigma.

· Based on scale created by Corrigan

· Average score is ,12.82 suggests average person in sample falls in middle of scale. Moderate stigma

· Reasonable reliability Cronbach’s alpha .837

Frequencies Independent Variables

CONTACT

· Six items asking about the extent to which the respondent has contact with the mentally ill. Questions ranged from seeing a mentally ill on an infrequent basis to living with a person with mental illness. Items were scored 1 if “yes” and 0 if “no.” Almost half the sample (47%) stated they have a relative who has mental illness and another 10% live with someone who is mentally ill.

· For the purposes of further analysis, respondents were given the score of the item with most extensive contact that they responded “yes” to. Possible scores ranged from 1 to 6. A higher score indicates greater contact.

Scale: LITERACY

· Five items asking about the extent to which an individual is knowledgeable about mental illness. Each item represents a myth about mental illness (e.g. Mental illness is not as common as cancer, diabetes or heart disease.) Items were either true or false and correct answers were scored “1.” The number of correct answers were compiled with the lowest possible score 0 and highest 5.

· Based on material developed by the WHO

· Average score is ,3.5 suggests average person in sample moderately knowledgable.

· Reliability poor (Cronbach’s alpha=.185) suggesting knowledge of one piece of information does not necessarily mean knowledge of another.

Scale: MEDIA

· Four items asking about the extent to which the respondent has noticed discussion of mental illness or treatment for mental illness in the media. Scored 1-5 with higher score indicating greater awareness. Possible scores 5 to 25. Higher score indicates greater stigma.

· Average score is 14.41 suggests average person in sample falls in middle of scale. Moderate awareness

· Weak reliability Cronbach’s alpha .439

DIAGNOSIS

Three vignettes were presented describing: Nancy, a depressed individual, Richard an individual with drinking problems and Harry, an individual who is schizophrenic. In order to measure the extent to which a diagnosis might be more or less stigmatized three questions were asked, based upon the work of Corrigan. The frequency distributions are below

Because of very small number in category, for analytic purposes 15 responding to Richard excluded from analysis along with missing

Because of very small number in category, for analytic purposes 4 people responding to Nancy excluded from analysis along with missing

Because of very small number in category, for analytic purposes 4 people responding to Nancy excluded from analysis along with missing

· Most sympathetic to individual with schiz. (over half the sample) Least sympathy for problem drinker.

· Most (42%) would avoid schiz. Negligible % would avoid depressed.

· 41% of sample felt schiz would most benefit from treatment – more than other diagnoses. One third thought depressed individual would most benefit from treatment

ETIOLOGY

In an attempt to measure etiology of mental illness after these vignettes questions asked about the causes of each illness. Possible answers included trauma, genes, chemical imbalance family and life stressors. For analytic purposes the responses were grouped into three categories in tables below.

· Almost two thirds of the sample responded depression caused by life stressors. One quarter biological

· Almost ¾ of sample responded problemn drinking caused by life stressors

· Almost two thirds responded schiz. rooted in biological causes

BIVARIATE ANALYSIS

All of the dependent variables are significantly correlated with one another. (Describe one by one using r and r2)

More stereotypes more distance less openness to TX and greater stigma

	ATTIT	DISTANCE	OPENTX	STIGMA
ATTIT	Pearson Correlation	1	-.322**	.336**	.347**
	Sig. (2-tailed)		.000	.000	.000
	N	183	183	179	179
DISTANCE	Pearson Correlation	-.322**	1	-.278**	-.219**
	Sig. (2-tailed)	.000		.000	.003
	N	183	183	179	179
OPENTX	Pearson Correlation	.336**	-.278**	1	.181*
	Sig. (2-tailed)	.000	.000		.015
	N	179	179	179	178
STIGMA	Pearson Correlation	.347**	-.219**	.181*	1

Variable	SPSS Variable Name	ATTIT	DISTANCE	OPEN	STIGMA
Contact with the mentally ill	CONTACT	r= -.16 p<.03	r= .27 p<.01	r= -.21 p<.01	r= -.13 p<.10
Mental Health Literacy	LITERACY	r=-.06 ns	r= -.24p<.01	r=-.06 ns	r=-.02 ns
Media Awareness	MEDIA	r= .01 ns	r=-.04 ns	r=-.11 ns	r=-.02 ns
Diagnosis Who do you have most sympathy for Who would you most avoid Who would most benefit from treatment	SYMP · Depressed · Schiz · None AVOID · Alcohol · Schiz · None BENEFIT · Depressed · Alcohol · Schiz	F= .609 ns F=7.27p <.01 26.4 26.56 20.21 F=.39 ns	F= 1.273 ns F=2.83p<.06 4.00 3.86 4.51 F=.75 ns	F=.723 ns F=1.41 ns F=1.88 ns	F =.931 ns F=5.95p<.01 11.85 14.09 11.09 F=1.15 ns
Etiology	DEPRCAUSE · Biological · Trauma · Life stressors ALCAUSE · Biological · Trauma · Life stressors SCZCAUSE · Biological · Trauma · Life stressors	F=.03 ns F=1.5 ns F=3.39 p<.04 24.57 24.52 31.17	F=.35 ns F=3.4 p<.04 4.00 4.74 3.92 F=1.29 ns	F=2.9 p<.06 10.76 10.58 12.34 F.=.51 ns F=2.56 p<.08 11.27 12.19 13.5	F=.44 ns F=.22 ns F=1.37 ns

Anything is bold is significant

Contact significantly related to three of four dependent variables and approaches significance with other. More contact less stereotypes less distance want more open to treatment and less stigma.

Mental health literacy related to distance; less distance more literate

Diagnosis: Those who would avoid none of the people described less stereotypes less distance. Those who would avoid schiz more stigma

Etiology: Those who believe depression caused by life stressors less open to tx; Those who believe problem drinking due to life stressors more distance; those who believe schiz caused by life stressors more stereotypical attitudes and less open to tx compared to those who believe caused by biology or trauma

Variable	SPSS Variable Name	ATTIT	DISTANCE	OPEN	STIGMA
Age	AGE	r=-.04 ns	r=-.09 ns	r= .01 ns	r04 ns
Gender	GENDER · Male · Female	t= 3.10 p<.002 29.28 23.66	t= -.695 ns	t = 3.57 <.001 13.35 10.88	t = .566 ns
Education	EDUC3 · <college · college grad · >college	F= 2.57 P<.07 26.74 26.91 22.75	F= 1.411 ns	F=3.96 p<.02 12.96 11.34 10.77	F = .634 ns
Ethnicity	WHITE · White · /nonwhite	t =.51 ns	t = -.15 ns	t = 1.45 ns	t = -1.78 p<.08 13.23 11.63
Income	INCOME3 · Less than $39,999 · $40,000-69,999 · 70,000+	F = .13 ns	F = 1.72 ns	F = .067 ns	F= .35 ns

· Means

Women fewer stereotypes and more open to tx

Whites more stigma than non-whites (approaches sign)