Descriptive Statistics
Mothers were on average 32.95 years old (SD = 4.63). Most of them were born in Germany (91.5 %), had a partner (97.8 %), and attained more than ten years of education (73.9 %). Detailed sociodemographic characteristics of the sample are presented in Table 1.
Table 1
Sociodemographic Characteristics of the Sample
|
M (SD)
|
Range
|
Maternal age a
|
32.95±4.63
|
16.8–54.0
|
Age of child b
|
13.23±2.73
|
6.0–25.6
|
|
n = 3,507
|
%
|
Duration of residence in Germany c
|
|
Born in Germany
|
3,210
|
91.5
|
<5 years
|
87
|
2.5
|
5–10 years
|
103
|
2.9
|
>10 years
|
109
|
3.1
|
Partnership status
|
|
Partner
|
3,432
|
97.8
|
No partner
|
77
|
2.2
|
Education
|
|
≤10 years
|
914
|
26.0
|
>10 years
|
2,592
|
73.9*
|
Net income d
|
|
<1,250 €
|
84
|
2.4
|
1,250 €–2,249 €
|
356
|
10.2
|
2,250 €–2,999 €
|
440
|
12.6
|
3,000 €–3,999 €
|
895
|
25.6
|
4,000 €–4,999 €
|
883
|
25.3
|
>5,000 €
|
837
|
24.0*
|
Number of children
|
|
1
|
1,843
|
52.6
|
2
|
1,275
|
36.4
|
3
|
298
|
8.5
|
4
|
67
|
1.9
|
≥5
|
22
|
0.6
|
IPV groups
|
|
|
Without
|
1,813
|
51.9
|
Psychological
|
941
|
26.9
|
Physical
|
278
|
7.9
|
Sexual
|
297
|
8.5
|
Sexual and physical
|
172
|
4,9
|
Note. n varied slightly due to missing values. IPV = Intimate Partner Violence.
a in years, b in weeks, c time since migration to Germany, d per month and household.
* the valid percentages do not add up to 100% due to rounding errors.
IPV Prevalence
In this sample, 48.2% of women disclosed instances of IPV. Specifically, 45.2% of women had encountered psychological IPV, with 13.5% experiencing it within the past 12 months. Additionally, 12.8% reported incidents of physical IPV, with 1.9% being affected in the last 12 months. Moreover, 13.3% of women were affected by sexual IPV, with 1.3% experiencing it within the past 12 months. As shown in Figure 2, most women affected by sexual and/or physical IPV also reported experiences of psychological IPV, for which the highest lifetime prevalence was found.
Preferences for Counseling and Treatment Services and Mode of Service Provision
Table 2 shows the descriptive statistics of preferences for all service and mode of service provision items across all groups. Among the various services, all women preferred the items “family member, friend, or colleague” and “woman in the same situation” the most and preferred “religious institutions” the least. “In person” was the most popular modality of service provision, followed by “telephone call” and “video conference”. Women rated “apps or online platforms without guidance from an expert” as the least popular.
Table 2
Descriptive Statistics of Preferences for Individual Counseling and Treatment Services as well as for Mode of Service Provision
Item
|
M (SD)
|
Likeliness of picking a service:
|
Counseling and treatment service preferences
|
Definitely
(4)
|
Family member, friend, or colleague
|
3.3 (0.8)
|
Rather yes
(3)
|
Woman in the same situation
|
3.2 (0.7)
|
Women's shelter
|
3.1 (0.8)
|
Intervention center
|
3.0 (0.7)
|
Emergency Room
|
3.0 (0.9)
|
Help hotline for violence against women or sexual abuse
|
2.9 (0.8)
|
Life and family counseling center
|
2.9 (0.7)
|
Counseling service for victims of crime
|
2.9 (0.8)
|
Family midwife
|
2.9 (0.9)
|
Midwife
|
2.8 (0.9)
|
Gynecologist
|
2.8 (0.9)
|
Outpatient clinic for psychiatry or psychosomatic
medicine
|
2.8 (0.8)
|
General Practitioner
|
2.7 (0.9)
|
Self-help group
|
2.6 (0.9)
|
Psychosocial crisis service of the health department
|
2.5 (0.8)
|
Telephone counseling
|
2.5 (0.8)
|
Social pedagogical family assistance
|
2.5 (0.8)
|
Rather no
(2)
|
Pediatrician
|
2.3 (1.0)
|
Religious institutions
|
1.7 (0.8)
|
|
Not at all
(1)
|
Mode of service provision preferences
|
Definitely
(4)
|
In person
|
3.7 (0.5)
|
|
Telephone call
|
2.7 (0.8)
|
Rather yes
(3)
|
Video conference
|
2.6 (0.8)
|
App or online platform with guidance from expert
|
1.9 (0.8)
|
Rather no
(2)
|
E-mail
|
1.8 (0.8)
|
Chat
|
1.8 (0.8)
|
App or online platform without guidance from expert
|
1.5 (0.6)
|
Not at all
|
|
|
(1)
|
Note. n = 3,507. n varied slightly due to missing values.
In addition to these services, women were asked if they would contact the police or any other law enforcement agency. In this sample, 2,522 women (71.9 %) affirmed that they would contact the police, while 728 women (20.7 %) answered “don´t know”, and 259 women (7.4 %) answered “no”.
Correlation Analysis
Spearman’s rank correlation was computed to assess the relationships between covariates and outcome variables. A higher score of counseling and treatment service preferences was statistically significantly associated with higher maternal age (r = .053, p = .001). A higher score of modes of service provision preferences was statistically significantly associated with a shorter duration of residence in Germany (r = -.044, p = .001) and a lower number of children (r = -.043, p = .05). Therefore, maternal age was included in the following analysis for service preferences (for the total score as well as for every service subscale). Likewise, duration of residence in Germany and number of children were included in the analysis for mode of service provision preferences. Detailed correlations can be found in Appendix 3.
Group Comparisons for Counseling and Treatment Service Preferences
After adjusting for maternal age and its interaction term due to unmet homogeneity requirement (maternal age*IPV groups), IPV groups differed significantly in their total scores for preferences for services [F(4,3489) = 3.161, p = .013, partial η² =.004]. Bonferroni-corrected post hoc tests revealed a lower total score for counseling and treatment service preferences for women affected by any type of IPV compared to women without experiences of IPV, indicating that they preferred all services less than non-affected women. Furthermore, the total score for preferred services of women affected by physical IPV or women affected by sexual IPV was significantly lower than that of women affected by psychological IPV (see Table 3). Neither maternal age as a covariate [F(1, 3489) = 1.931, p = .165] nor its interaction term [F(1, 3489) = 2.741, p =.098] showed statistical significance. Considering group comparisons for the three subscales of service preferences, 15 multivariate outliers were identified and excluded from further calculations. After adjusting for maternal age, one-way MANCOVA with bootstrapping revealed differences between groups and the subscales [F(12, 9186.340) = 5.806, p <.001, partial η² =.007, Wilk’s Λ =.980]. Moreover, maternal age as a covariate showed significance [F(3, 3472.000) = 12.029, p <.001, partial η² =.010, Wilk’s Λ =.990]. Post hoc univariate ANOVAs were conducted for the three subscales and indicated statistically significant differences between IPV groups in every subscale, namely preferences for psychosocial services [F(4, 3474) = 11.105, p <.001, partial η² =.013], medical services [F(4, 3474) = 2.665, p =.031, partial η² =.003], and midwives [F(4, 3474) = 10.755, p <.001, partial η² =.012]. Bonferroni-corrected post hoc analyses with bootstrapping revealed that psychosocial services were less preferred by women affected by any type of IPV compared to non-affected women. Furthermore, women affected by physical IPV and women affected by sexual IPV reported that they would be less likely to use psychosocial services than women affected by psychological IPV. Detailed information can be found in Fehler! Verweisquelle konnte nicht gefunden werden.3. Medical services were rated significantly less likely to be used by women affected by psychological IPV and women affected by physical IPV than by non-affected women. In this subscale, higher maternal age was additionally associated with significantly higher ratings for medical service preferences [F(1, 3474) = 26.678, p <.001, partial η² =.008]. As shown in Fehler! Verweisquelle konnte nicht gefunden werden.3, midwives were significantly less likely to be used by women affected by any type of IPV compared to non-affected women, and women affected by physical or sexual IPV rated midwives less likely to be used than women affected by psychological IPV. According to Cohen, the effect sizes were small for the subscales of psychosocial service preferences (partial η² = .013) and preferences for midwives (partial η² = .012) and very small for the subscale of medical service preferences (partial η² = .003).
Table 3
Group Comparisons for Counseling and Treatment Service Preferences
|
IPV group (I)
|
IPV group (J)
|
Services
|
Psychosocial services
|
Medical services
|
Midwives
|
M (I─J)
|
P
|
95 % BCa CI
|
M (I─J)
|
p
|
95 % BCa CI
|
M
(I─J)
|
p
|
95 % BCa CI
|
M (I─J)
|
p
|
95 % BCa CI
|
LL
|
UL
|
LL
|
UL
|
LL
|
UL
|
LL
|
UL
|
Without
|
Psychological
|
2.06*
|
.015
|
0.53
|
3.73
|
0.57**
|
.004
|
0.22
|
0.96
|
0.27**
|
.016
|
0.06
|
0.48
|
0.20**
|
.002
|
0.08
|
0.33
|
Physical
|
4.42**
|
.007
|
1.51
|
7.94
|
1.44***
|
<.001
|
0.84
|
2.06
|
0.39*
|
.032
|
0.02
|
0.74
|
0.43***
|
<.001
|
0.23
|
0.63
|
Sexual
|
5.32*
|
.023
|
1.10
|
10.37
|
1.35***
|
<.001
|
0.76
|
1.95
|
0.28
|
.092
|
-0.05
|
0.63
|
0.47***
|
<.001
|
0.26
|
0.66
|
Sexual and physical
|
6.28*
|
.040
|
0.81
|
12.71
|
1.25**
|
.008
|
0.39
|
2.13
|
0.32
|
.170
|
-0.12
|
0.80
|
0.38***
|
<.001
|
0.14
|
0.61
|
Psychological
|
Without
|
-2.06*
|
.015
|
-3.73
|
-0.53
|
-0.57**
|
.004
|
-0.96
|
-0.22
|
‑0.27**
|
.016
|
-0.48
|
‑0.06
|
-0.20**
|
.002
|
‑0.33
|
‑0.08
|
Physical
|
2.37**
|
.008
|
0.79
|
4.19
|
0.87*
|
.011
|
0.23
|
1.56
|
0.12
|
.540
|
-0.28
|
0.48
|
0.22*
|
.039
|
0.02
|
0.43
|
Sexual
|
3.26*
|
.046
|
0.33
|
6.68
|
0.77*
|
.014
|
0.12
|
1.44
|
0.01
|
.943
|
‑0.35
|
0.37
|
0.26*
|
.018
|
0.03
|
0.47
|
|
Sexual and physical
|
4.22
|
.065
|
0.18
|
9.04
|
0.68
|
.148
|
-0.21
|
1.62
|
0.05
|
.830
|
-0.45
|
0.55
|
0.18
|
.160
|
-0.09
|
0.43
|
Physical
|
Without
|
‑4.42**
|
.007
|
-7.94
|
-1.51
|
-1.44***
|
<.001
|
-2.06
|
-0.84
|
-0.39*
|
.032
|
-0.74
|
-0.02
|
‑0.43***
|
<.001
|
-0.63
|
-0.23
|
Psychological
|
-2.37**
|
.008
|
-4.19
|
-0.79
|
-0.87*
|
.011
|
-1.56
|
-0.23
|
-0.12
|
.540
|
-0.48
|
0.28
|
-0.22*
|
.039
|
-0.43
|
‑0.02
|
Sexual
|
0.90
|
.342
|
-0.95
|
2.99
|
-0.10
|
.808
|
-0.95
|
0.64
|
-0.10
|
.642
|
-0.58
|
0.35
|
0.04
|
.760
|
-0.24
|
0.29
|
Sexual and physical
|
1.86
|
.238
|
-1.12
|
5.30
|
-0.19
|
.709
|
-1.28
|
0.82
|
-0.07
|
.836
|
-0.60
|
0.49
|
-0.05
|
.736
|
-0.34
|
0.22
|
Sexual
|
Without
|
-5.32*
|
.023
|
‑10.37
|
-1.10
|
-1.35***
|
<.001
|
-1.95
|
-0.76
|
-0.28
|
.092
|
-0.63
|
0.05
|
-0.47***
|
<.001
|
-0.66
|
-0.26
|
Psychological
|
-3.26*
|
.046
|
-6.68
|
-0.33
|
-0.77*
|
.014
|
-1.44
|
-0.12
|
-0.01
|
.943
|
-0.37
|
0.35
|
-0.26*
|
.018
|
-0.47
|
-0.03
|
Physical
|
-0.82
|
.342
|
-2.99
|
0.95
|
0.10
|
.808
|
-0.64
|
0.95
|
0.10
|
.642
|
-0.35
|
0.58
|
-0.04
|
.760
|
-0.29
|
0.24
|
|
Sexual and physical
|
0.96
|
.321
|
-0.93
|
3.21
|
-0.10
|
.859
|
-1.04
|
0.95
|
0.04
|
.862
|
-0.46
|
0.59
|
-0.09
|
.570
|
-0.37
|
0.21
|
Sexual and Physical
|
Without
|
-6.28*
|
.040
|
-12.71
|
-0.81
|
-1.30**
|
.008
|
-2.13
|
-.39
|
-0.32
|
.170
|
-0.80
|
0.13
|
-0.38***
|
<.001
|
-0.61
|
-0.14
|
Psychological
|
-4.22
|
.065
|
-9.04
|
-0.18
|
-0.68
|
.148
|
-1.62
|
0.21
|
-0.05
|
.830
|
-0.55
|
0.45
|
-0.18
|
.160
|
-0.43
|
0.09
|
Physical
|
-1.86
|
.238
|
-5.30
|
1.12
|
0.19
|
.709
|
-0.82
|
1.28
|
0.07
|
.836
|
-0.49
|
0.60
|
0.05
|
.736
|
-0.22
|
0.34
|
Sexual
|
-0.96
|
.321
|
-3.21
|
0.93
|
0.10
|
.859
|
-0.95
|
1.04
|
-0.04
|
.862
|
-0.59
|
0.46
|
0.09
|
.570
|
-0.21
|
0.37
|
Note. IPV = Intimate Partner Violence. Bootstrap results are based on 1,000 bootstrap samples. BCa CI = bias-corrected and accelerated bootstrap interval, LL = lower limit, UL = upper limit. Calculations without outliers and with covariate: maternal age.
*p < .05. **p < .01. ***p < .001.
|
Group Comparisons for Mode of Service Provision Preferences
Considering mode of service provision preferences, one univariate extreme outlier was found and excluded from the calculations. IPV groups did not differ significantly in their preferences for modes of service provision in the total score [F(4,3491) = 1.290, p = .272, partial η² = .001] after adjusting for number of children and duration of residence in Germany. Both covariates, however, showed statistical significance: a longer duration of residence in Germany [F(1, 3491) = 9.800, p =.002, partial η² = .003] and a higher number of children [F(1, 3491) = 11.132, p =.001, partial η² = .003] were associated with lower scores in preferences for mode of service provision.
Considering group comparisons for the two subscales of mode of service provision preferences, 24 outliers were found and excluded. Groups differed on the subscales as revealed with a one-way MANCOVA [F(8, 6932.00) = 2.569, p =.009, partial η² =.003, Wilk’s Λ = .994)]. Duration of residence in Germany [F(2, 3466.000) = 22.588, p <.001, partial η² =.013, Wilk’s Λ =.987] and number of children [F(2, 3466.000) = 7.526, p <.001, partial η² =.004, Wilk’s Λ =.996)] were also statistically significant in this model. Post hoc univariate ANOVAs showed a statistically significant difference between IPV groups for the subscale of direct modes of service provision [F(4, 3467) = 4.577, p =.001, partial η² =.005] but not for the subscale of indirect modes of service provision [F(4, 3467) =.259, p =.904, partial η² <.001]. The covariate number of children showed statistical significance for the subscale of direct modes of service provision [F(1, 3467) = 14.222, p <.001, partial η² =.004] and the covariate duration of residence in Germany for the subscale of indirect modes of service provision [F(1, 3467) = 32.533, p < .001, partial η² =.009]. Higher values of the covariates were associated with lower scores for preferences in the dependent variables. Bonferroni-corrected post hoc tests with bootstrapping for the subscale direct modes of service provision revealed that women affected by physical and/or sexual IPV rated direct modes less favorably than women without experiences of IPV (see Table 4). Likewise, women affected by physical IPV and/or sexual IPV rated direct modes less favorably than women affected by psychological IPV.
Table 4
Group Comparisons for the Subscale Direct Modes of Service Provision Preferences
IPV group (I)
|
IPV group (J)
|
M (I─J)
|
p
|
95 % BCa CI
|
LL
|
UL
|
Without
|
Psychological
|
0.00
|
.903
|
-0.04
|
0.04
|
Physical
|
0.10**
|
.005
|
0.03
|
0.16
|
Sexual
|
0.07*
|
.022
|
0.01
|
0.13
|
Sexual and physical
|
0.10*
|
.017
|
0.02
|
0.19
|
Psychological
|
Without
|
-0.00
|
.903
|
-0.4
|
0.04
|
Physical
|
0.09*
|
.010
|
0.02
|
0.17
|
Sexual
|
0.07*
|
.032
|
0.00
|
0.13
|
|
Sexual and physical
|
0.10*
|
.019
|
0.01
|
0.19
|
Physical
|
Without
|
-0.10**
|
.005
|
-0.16
|
-0.03
|
Psychological
|
-0.09*
|
.010
|
-0.17
|
-0.02
|
Sexual
|
-0.03
|
.474
|
-0.12
|
0.06
|
Sexual and physical
|
0.01
|
.898
|
-0.10
|
0.12
|
Sexual
|
Without
|
-0.07*
|
.022
|
-0.13
|
-0.01
|
Psychological
|
-0.07*
|
.032
|
-0.13
|
-0.00
|
Physical
|
0.03
|
.474
|
-0.06
|
0.12
|
|
Sexual and physical
|
0.04
|
.494
|
-0.07
|
0.14
|
Sexual and Physical
|
Without
|
-0.10**
|
.017
|
-0.19
|
-0.02
|
|
Psychological
|
-0.10**
|
.019
|
-0.19
|
-0.01
|
|
Physical
|
-0.01
|
.898
|
-0.12
|
0.10
|
|
Sexual
|
-0.04
|
.494
|
-0.14
|
0.07
|
Note. IPV = Intimate Partner Violence. Bootstrap results are based on 1,000 bootstrap samples. BCa CI = bias-corrected and accelerated bootstrap interval, LL = lower limit, UL = upper limit. Calculations without outliers and with covariates: duration of residence in Germany, number of children.
*p < .05. **p < .01.