Beskæftigelsesudvalget 2023-24
BEU Alm.del Bilag 127
Offentligt
2843542_0001.png
1228158
research-article2024
sjP0010.1177/14034948241228158M.B.D. Nielsen et al.Two Methods for Assessing Workplace sexual and Gender-based harassment
Beskæftigelsesudvalget 2023-24
BEU Alm.del - Bilag 127
Offentligt
Scandinavian Journal of Public Health,
1–8
OriginAl ArTicle
Workplace sexual and gender-based harassment in Denmark: a
comparison of the self-labelling and behavioural list method
MAj BrITT DAhl NIelseN
1
, AlBerTe BOukAIDI ANDerseN
2
,
Gry GruNDTvIG
3
, kAThrINe søreNseN
2,4
, jOsefINe rANfelT ANDerseN
1
,
NANNA P. lArssON
2
, sOfIe sMeDeGAArD skOv
1
,
ANNA PAlDAM fOlker
1
, susIe kjær
5
, Per TyBjerG AlDrIch
6
,
reINer ruGulIes
2,7
, ThOMAs clAuseN
2
& IDA e. h. MADseN
1,2
National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark,
2
The National Research
Centre for the Working Environment, Copenhagen, Denmark,
3
COWI, Lyngby, Denmark,
4
Department of Psychology,
University of Copenhagen, Copenhagen, Denmark,
5
Anerkendende Psykologpraksis, Herlev, Denmark,
6
BUPL Copenhagen
Ø, Denmark,
7
Section of Epidemiology, Department of Public Health, University of Copenhagen, Copenhagen, Denmark
1
The
Abstract
Aim:
knowledge about the prevalence of sexual and gender-based harassment is hampered by disagreements about definitions
and measurement methods. The two most common measurement methods are the self-labelling (a single question about
exposure to sexual harassment) and the behavioural list method (an inventory of sexually harassing behaviours). The aim of
this paper was to compare the self-labelling and the behavioural list methods for measuring sexual harassment and assess the
association with depressive symptoms.
Methods:
The study is based on a convenience sample of 1686 individuals employed in
29 workplaces in Denmark. survey data were collected from November 2020 until june 2021 and there were 1000 participants
with full data on key variables. We used a linear mixed-effects model to examine the relationship between sexual harassment
and depressive symptoms.
Results:
In total, 2.5% self-labelled as being sexually harassed, while 19.0% reported exposure to at
least one type of sexual and gender-based harassment using the behavioural list method. Both groups reported higher levels
of depressive symptoms compared with non-exposed employees. The most common types of behaviours were: that someone
spoke derogatorily about women/men (11.6%); being belittled because of one’s gender or sexuality (4.7%); and unwanted
comments about one’s body, clothes or lifestyle (4.5%).
Conclusions:
The behavioural list method yielded a higher
prevalence of sexual and gender-based harassment compared with the self-labelling method. Self-labelling and
reporting at least one type of sexual and gender-based harassment was associated with depressive symptoms.
Keywords:
sexual harassment, self-labelling, survey, prevalence, depressive symptoms
Background
Workplace sexual harassment is associated with
depression, anxiety, post-traumatic stress disorder
and suicide[1-7]. There is no universally accepted
definition, but most highlight that it is unwanted and
harms the victim and/or contributes to a hostile
working environment [2,8-10]. One of the most
widely used conceptual frameworks is the Tripartite
Model [2] that defines sexual harassment as three
distinct, but interrelated phenomena: unwanted sex-
ual attention; coercion; and gender harassment.
Unwanted sexual attention
encompasses unrecipro-
cated sexual advances,
coercion
includes job-related
pressures or bribes, for example, a promotion in
exchange for sexual favours or threats of employment
correspondence: Maj Britt Dahl Nielsen, The National Institute of Public health, university of southern Denmark, studiestræde 6, Dk-copenhagen,
1455, Denmark. e-mail: [email protected]
Date received 30 September 2022; reviewed 13 December 2023; accepted 7 January 2024
© Author(s) 2024
Article reuse guidelines: sagepub.com/journals-permissions
https://doi.org/10.1177/14034948241228158
DOI: 10.1177/14034948241228158
journals.sagepub.com/home/sjp
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
2
M.B.D. Nielsen et al.
harassment [16]. seQ was originally developed to
measure male-to-female harassment, and although it
has been used in different countries, it is not a stand-
ardized survey, as the number of items and wordings
differs [17]. seQ was developed 30 years ago, and
while sexual harassment may not have changed fun-
damentally, major changes in the way we work, for
example, due to digitalization, may not be adequately
reflected [18,19]. Moreover, other surveys have been
developed specially for the Nordic and european
context, including the Bergen sexual harassment
scale (Bshs) [4,20] and the survey of the european
union Agency for fundamental rights (frA) [18].
The latter only focused on women and was not
designed to measure workplace sexual harassment,
and neither Bshs nor frA include non-sexual gen-
der harassment.
We aimed to develop an inventory covering a broad
range of behaviours that may constitute sexual and
gender-based harassment to aid preventive efforts. We
developed the survey in collaboration with workplaces
and labour market stakeholders to strengthen its rel-
evance and acceptability to support a more uniform
way of measuring sexual and gender-based harass-
ment in Denmark. This is particularly important
given the controversies surrounding the definition
[2]. here, we explain the three phases of the process.
Phase 1: Theoretical and conceptual framework.
IWs is
rooted in a broad understanding of sexual harass-
ment inspired by fitzgerald’s Tripartite Model and
Berdahl’s conceptualization of sexual harassment.
According to Berdahl [21], sexual harassment is pri-
marily driven by the motivation to protect one’s sex
status, rather than sexual desire. Thus, it can be
understood as a punitive mean of
doing gender,
which
moves focus to a broader range of behaviours, that is,
sex-based slurs and exclusion [21]. It challenges the
traditional notion that only men harass women,
because the desire to protect one’s sex status is held
by men and women alike [21].
Phase 2: Item development.
We reviewed different sur-
veys and discussed their strengths and weaknesses.
Next, we developed a first draft of IWs inspired by a
Danish version of the sexual experience Question-
naire–Department of Defence (seQ-DoD), the sex-
ual harassment Inventory (shI), Bshs and frA,
and the theoretical and conceptual framework
described in the previous section. Table I shows over-
laps between IWs and surveys used for inspiration.
We discussed the first draft with an advisory board
consisting of key labour market stakeholders (about
20 unions, employer organizations and researchers)
and conducted a parallel qualitative study among
employees exposed to sexual harassment.
if demands are not met, and
gender harassment
includes hostile or degrading behaviours and atti-
tudes about another person’s gender or sex [10,11].
knowledge about the prevalence is severely ham-
pered by disagreements about definitions and meas-
urement methods. Thus, researchers, policy makers
and workplaces have no standards against which to
measure progress in reducing sexual harassment [12].
The two most common measurement methods are the
self-labelling method (typically a single-item question
asking the respondents whether they have experienced
sexual harassment) and the behavioural list method
(an inventory of sexually harassing behaviours). When
using the behavioural list method, the respondent
does not have to label these behaviours as sexual har-
assment, and the method is therefore not affected by
the respondent’s subjective definition of sexual harass-
ment. Previous research suggests that the behavioural
list method yields higher prevalence compared with
the self-labelling method [13]. Possible reasons for the
self-labelling method underestimating the prevalence
of sexual harassment include: sexual harassment being
associated with considerable social stigma, that is,
beliefs that targets of it are complainers, weak and
powerless [14], and people may reject this label to
avoid social stereotypes and maintain positive self-
images [14]. Moreover, respondents may be unsure
whether their experience constitutes sexual harass-
ment without pre-established definitions.
National surveys on sexual harassment in Denmark
and other scandinavian countries have been based on
the self-labelling method. In 2018, 5.5% of employed
women and 1.8% of employed men in Denmark
reported exposure to sexual harassment in the previ-
ous 12 months [15]. While these sex differences could
reflect that women are more often targets of sexual
harassment, they could also reflect sex differences in
self-labelling, for example, because sexual harassment
is often framed as a women’s problem. The aim of this
research project was to measure workplace sexual and
gender-based harassment in Denmark using the
behavioural list method. We developed a new survey,
the Inventory of Workplace sexual and Gender-based
harassment (IWs). In this article, we describe the
development process, compare the prevalence using
IWs and the self-labelling method in a sample of
Danish employees, and examine the cross-sectional
association between sexual harassment and self-
reported depressive symptoms.
Methods
Designing IWS
One of the most widely used surveys is the sexual
experience Questionnaire (seQ), which measures
unwanted sexual attention, coercion and gender
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
2843542_0003.png
Two Methods For Assessing Workplace Sexual and Gender-based Harassment
Table I. Prevalence of sexual harassment according to self-labelling and the IWs.
Number
of persons
Have you been sexually harassed in your workplace during the last 12 months? (Total)
Men
Women
Persons reporting at least one of the 21 iWS items, total
Men
Women
How many times during the last 12 months have you been exposed to the following in connection with your work:
(1) unwanted sexual comments about your body, clothes or lifestyle
a,b,c,d
(2) unwanted sexual comments in a larger group/gathering
b
(3) unwanted messages with sexual content (e.g. letters, text messages, emails or messages on social media)
a,d
(4) staring or flirtatious glances with sexual undertones that were unwanted or unpleasant to you
a,b,c,d
(5) unwanted physical contact with sexual undertones (e.g. pat, kiss or hug)
a,b,c,d
(6) unwanted movements with sexual undertones directed at you
b
(7) unsolicited requests for dates even if you have already said no
b,c,d
(8) someone showing you pornographic images or other material with sexual content that were unwanted or
unpleasant to you
a,b,c,d
(9) someone whistling at or catcalling you in a way that was unwanted or unpleasant to you
b,c
(10) someone exposing themselves to you (e.g. taken off their clothes) in a way that was unwanted or uncomfortable for
you
b,c,d
(11) someone bringing you into conversations about sex against your will
b
(12) someone spreading sexual rumours about you
a,b
(13) someone telling stories with sexual content that were unwanted or unpleasant to you
b
(14) someone speaking derogatorily about women/men in a way that was unwanted or unpleasant to you
b,c
(15) sexually explicit activities (e.g. games or strippers) at festive events that were unwanted or unpleasant to you
c
(16) someone belittling you because of your gender or sexuality
b,c
(17) someone excluding you from social gatherings or social network because of your gender or sexuality
b
(18) someone asking you for sexual favours in exchange for a reward (e.g. a pay rise or promotion)
a,b,c
(19) someone threatening punishment or sanctions (e.g. firing) if you refused their requests for sexual favours
a,b,c
(20) someone touching you against your will (e.g. groped you and/or held on to you)
(21) Attempted rape or actual rape
a,b,c
a
Bshs[4,20].
b
seQ-DoD[28,29].
c
shI[30].
d
frA[18].
3
Percentage
share
2.5%
1.0%
4.3%
19.0%
14.8%
24.3%
4.5%
3.5%
2.6%
2.3%
1.6%
0.6%
0.7%
1.7%
0.5%
0.6%
2%
0.8%
2.9%
11.6%
0.5%
4.7%
1.1%
0%
0%
0.6%
0%
25
6
19
190
83
105
45
35
26
23
16
6
7
17
5
6
20
8
29
116
5
47
11
0
0
6
0
Bshs: Bergen sexual harassment scale; frA: Agency for fundamental rights; IWs: Inventory of Workplace sexual and Gender-based harassment; shI:
sexual harassment Inventory; seQ-DoD: sexual experience Questionnaire–Department of Defence.
Phase 3: Field testing.
We conducted cognitive inter-
views with employees (some had experiences with
sexual harassment, others not) to ensure that ques-
tions were understandable and relevant. Besides
minor adjustments in wording, these interviews con-
tributed to the development of item 20, ‘someone
touching you against your will (e.g. groped you and/
or held on to you)’. This item was introduced to
encompass physical advances that do not constitute
rape or rape attempts but are more severe than
unwanted physical contact (item 6). finally, we dis-
tributed the survey in 29 workplaces and received
feedback from employers and employees.
IWs consists of two sections. The first section
measures sexual and gender-based harassment dur-
ing the last 12 months (Table I). In total, 11 items
cover different types of verbal harassment, 3 items
cover physical advances and 7 items cover non-verbal
behaviours. Item 3 covers online harassment, and
items 18 and 19 cover coercive behaviours. We also
included harassment based on the target’s sexual ori-
entation in item 16 and 17 because of the strong
theoretical link between them and the input from
stakeholders [22,23]. The second section includes
questions about workplace management and witness
experiences to aid management.
Data collection
The study is based on data from a convenience sam-
ple encompassing 29 workplaces recruited via the
researchers and the Advisory Boards Network. We
aimed to recruit a diverse group of workplaces, and
the sample consist of workplaces from the public and
private sector and from different industries and sizes.
In some workplaces, all employees received the sur-
vey, while in other workplaces only selected depart-
ments were invited (this was decided by the
management). The mean number of invited partici-
pants in each work unit was 57.5 individuals (10th
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
4
M.B.D. Nielsen et al.
yielding a scale from 0 to 50, with higher scores indi-
cating a higher level of depressive symptoms.
Workplace sexual harassment
was measured in two
ways. first, with the self-labelling method, using the
item from the national survey ‘Work environment
and health in Denmark’ [5,15] with the wording,
‘have you been sexually harassed in your workplace
in the last 12 months?’. response was ‘yes, daily’,
‘yes, weekly’, ‘yes, monthly’, ‘yes, rarely’ or ‘Never’.
No definition of sexual harassment was given in the
questionnaire. second, sexual harassment was meas-
ured using the behavioural list method with IWs
which consists of 21 items (see Table II). The
responses were recorded on a scale ranging from 1
‘Never’, 2 ‘Once’, 3 ‘Two–five times’ and 4 ‘More
than five times’. To examine the frequency, we cre-
ated a compiled binary IWs variable based on
whether the employees had experienced sexual har-
assment (⩾1) or none (0) of the IWs questions.
furthermore, we created a categorical variable for
sexual harassment, distinguishing between (a)
respondents who self-label as targets of sexual har-
assment, (b) respondents who report one or more
items from the IWs but do not self-label, and (c)
respondents who do not self-label and report none of
the IWs behaviours (non-exposed employees).
Statistical analyses
We used a linear mixed-effects model to examine the
relationship between sexual harassment and depres-
sive symptoms comparing the level/degree of depres-
sive symptoms in groups with and without exposure
to harassment. We conducted analyses comparing
levels of depressive symptoms between employees
that either reported sexual harassment measured by
self-labelling or reported sexual harassment meas-
ured by IWs (but did not self-label as targets of sex-
ual harassment), to employees who did not report
sexual harassment according to either method. We
also examined the association of each of the items
from the IWs with depressive symptoms.
We performed linear mixed-effects models in sAs
9.4 (sAs Institute Inc., cary, Nc) using PrOc
MIXeD to assess the association between sexual
harassment as a fixed effect (two levels: yes/No) and
the response variable (MDI score). By using a mixed
model with random intercepts across workplaces, we
accounted for correlated observations from partici-
pants being employed in the same workplace.
Assumptions of normality and homogeneity of vari-
ance of residuals was inspected by quantile–quantile
plots and residual plots. Due to differences in age,
gender and educational level across the different
workplaces, we decided to adjust for these three
percentile:
n
=
4, 90th percentile:
n
=
126). survey
data were collected from November 2020 until june
2021. A representative from the workplace sent a list
of email addresses to the researchers, who forwarded
an invitation letter by email with an online link to the
questionnaire to the employees. The letter included
information explaining participation in the survey
was voluntary, the purpose of the study and use of
data. A reminder was later sent. The management
received a report with results and was offered a con-
sultation with the research team to discuss findings.
In a few workplaces, the researchers presented the
results for the employees. Management never
received the data to ensure respondents anonymity.
In Denmark, register and questionnaire studies do
not require approval by committees on biomedical
research ethics according to Danish legislation. A
total of 1668 individuals were invited to respond to
the survey, and 1029 employees chose to do so, yield-
ing a 62% response rate. We excluded respondents
with missing key data (self-labelled sexual harass-
ment, IWs, Major Depression Inventory,
n
=
29),
yielding a final study population of 1000 participants
from 29 work units.
Depressive symptoms
were measured with the Major
Depression Inventory (MDI), an instrument that
allows measurement of depressive symptoms on a
continuous scale and to ascertain incidence of preva-
lent depressive disorders. The MDI has been com-
prehensively validated, particularly in Denmark.
Bech et al. tested the sensitivity and specificity of the
MDI against the Present state examination among
patients in psychiatric departments and among
healthy controls [24]. Olsen et al. examined the inter-
nal and external validity of the MDI among 91 psy-
chiatric patients [25]. Bech et al. evaluated the
standardization of the MDI as a depression severity
scale in a re-analysis of two previous studies, using
the visual Analogue scale as an index of external
validity [26]. Olsen et al. used the MDI to measure
depressive symptoms and prevalent depressive disor-
ders in the Danish general population [27]. These
studies concluded that the MDI is a reliable and valid
instrument for measuring the level of depressive
symptoms and identifying individuals with a proba-
ble depressive disorder. The MDI consists of 12
items rating symptoms of depression according to
the International classification of Diseases (IcD)-
10. The items were answered on a scale ranging from
0 ‘At no time’, 1 ‘some of the time’, 2 ‘slightly less
than half of time’, 3 ‘slightly more than half the time’,
4 ‘Most of the time’ to 5 ‘All of the time’. for two
pairs of items (items 8 and 9, and items 10 and 11)
only the item with the highest score was included in
the score. responses were scored 0–5 and summed,
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
2843542_0005.png
Two Methods For Assessing Workplace Sexual and Gender-based Harassment
Table II. Descriptive statistics for main study variables for the full
population.
study population
N
=
1000
gender
Women
Men
Age groups
18–25 years
26–35 years
36–45 years
46–55 years
55+ years
educational background
long-cycle higher education (Masters or equivalent)
Medium-cycle higher education (Bachelor or
equivalent)
short-cycle higher education
vocational school
high school
Other (primary school with or without graduation,
other)
Job category
skilled manual worker
Non-manual worker (e.g. clerk)
Manager
student
Apprentices
Other (self-employed, unskilled worker, supported
employment)
435 (43.5%)
562 (56.2%)
39 (3.9%)
268 (26.8%)
265 (26.5%)
248 (24.8%)
180 (18.0%)
465 (46.5%)
243 (24.30%)
100 (10.0%)
51 (5.1%)
70 (7.0%)
71 (7.1%)
5
56 (5.6%)
633 (63.3%)
204 (20.4%)
50 (5.0%)
15 (1.5%)
42 (4.2%)
As shown in Table II, 43% of the respondents were
women and 47% had a higher education.
Model 1 in Table III shows the estimated mean
difference in depressive symptoms between employ-
ees who self-labelled as being sexually harassed com-
pared with non-exposed respondents, and model 2
shows the estimated mean difference between
respondents exposed to at least one of the items in
IWs compared with non-exposed respondents.
employees who self-labelled as being sexually har-
assed had a higher mean level of depressive symp-
toms compared with the non-exposed, as did
employees who reported exposure to at least one
item in IWs but did not self-label.
Table Iv shows that exposure to six of the items in
IWs was statistically significantly associated with
higher levels of depressive symptoms, including item
1 (unwanted sexual comments about your body,
clothes or lifestyle), item 4 (unwanted physical
advances), item 5 (staring and flirtatious glances
with sexual undertone), item 14 (some talking nega-
tively about men/women) and items 16 and 17
(being belittled or excluded because of one’s gender
or sexuality).
Discussion
variables (fixed effects), as these factors are related to
depressive symptoms and workplace sexual harass-
ment. Information on age and gender of the respond-
ents was collected in the questionnaire. All analyses
used a level of statistical significance of
p
<
0.05, and
p
values were calculated using restricted Maximum
likelihood estimation. Model estimates are pre-
sented with confidence intervals (cIs) and
p
values.
We assessed the internal consistency of the IWs as a
scale, by calculating the cronbach’s alpha across the
21 items.
results
Table I shows the prevalence of sexual harassment,
and Table II shows the characteristics of the
respondents.
As shown in Table I, 25 respondents (2.5%) self-
labelled as being sexually harassed during the past 12
months (1.0% among men and 4.3% among women),
while 190 (19.0%) reported exposure to at least one
of the behaviours in IWs (14.8% among men and
24.1% among women). The respondents most often
reported that someone spoke derogatorily about
women/men (11.6%), belittled them because of their
gender or sexuality (4.7%), or exposed them to
unwanted comments about their body, clothes, or
lifestyle (4.5%). The cronbach’s alpha across all 21
IWs items was 0.73.
We found a higher prevalence of sexual harassment
using IWs (19.0%) compared with the self-labelling
method (2.5%). Previously, Nielsen et al. reported a
prevalence of 18.4% when using the Bshs [20],
consisting of 11 items and measures experienced
during the past 6 months. The frA survey showed
that one in five (21%) women in the european union
(eu) have been exposed to sexual harassment in the
previous 12 months. The frA survey used an
11-item inventory, and although the study only
included women and was not limited to sexual har-
assment at work, the prevalence is similar to the prev-
alence in our study [18]. Thus, it seems that the
behavioural list method produces relatively stable
prevalence, even though the wording and number of
items differ. When we calculated the internal consist-
ency of the IWs as a scale, the cronbach’s alpha was
0.7, indicating acceptable validity. however, we do
not assume that IWs measures one latent construct,
and more sophisticated analyses, like principal com-
ponent analyses or factor analyses, are needed to
fully assess the psychometric properties of the inven-
tory. We are currently conducting such analyses
based on a larger dataset that has recently been col-
lected. results are as yet unpublished.
Our findings support previous research showing
that sexual harassment is related to poor mental health
[1,4,5,7]. We found higher depressive symptoms
among employees self-labelling as being sexually
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
2843542_0006.png
6
M.B.D. Nielsen et al.
estimated mean difference
95% cI low
95% cI high
p
value
0.0145*
<0.0001*
Table III. Mean level of depressive symptoms in relation to sexual harassment recorded by self-labelling and IWs.
Association of sexual harassment and depressive symptoms
(N
=
1000)
employees who self-label as being sexually harassed compared with
3.42
0.68
6.16
non-exposed employees
employees exposed to at least one item in IWs but do not self-label
2.59
1.44
3.73
compared with non-exposed employees
Association of sexual harassment and depressive symptoms by gender (effect of at least one incident according to iWS)
Men exposed to at least one item in IWs regardless of self-labelling
2.39
0.89
3.89
compared with non-exposed employees
Women exposed to at least one item in IWs regardless of self-labelling 3.17
1.53
4.81
compared with non-exposed employees
0.0019
0.0002
All estimates are adjusted for gender and age, and educational level was added as a covariate. results are presented as estimates and 95% cI and *p
<
0.05
as significant.
cI: confidence interval; IWs: Inventory of Workplace sexual and Gender-based harassment.
Table Iv. Mean level of depressive symptoms in relation to sexual harassment according to IWs.
Association of the 21 IWs items with depressive symptoms (N
=
1000)
(1) unwanted sexual comments about your body, clothes or lifestyle
(2) unwanted sexual comments in a larger group/gathering
(3) unwanted messages with sexual content (e.g. letters, text messages, emails or
messages on social media)
(4) staring or flirtatious glances with sexual undertones that were unwanted or
unpleasant to you
(5) unwanted physical contact with sexual undertones (e.g. pat, kiss or hug)
(6) unwanted movements with sexual undertones directed at you
(7) unsolicited requests for dates even if you have already said no
(8) someone showing you pornographic images or other material with sexual
content that were unwanted or unpleasant to you
(9) someone whistling at or catcalling you in a way that was unwanted or unpleasant
to you
b,c
(10) someone exposing themselves to you (e.g. taken off their clothes) in a way that
was unwanted or uncomfortable for you
(11) someone bringing you into conversations about sex against your will
(12) someone spreading sexual rumours about you
(13) someone telling stories with sexual content that were unwanted or unpleasant
to you
(14) someone speaking derogatorily about women/men in a way that was unwanted
or unpleasant to you
(15) sexually explicit activities (e.g. games or strippers) at festive events that were
unwanted or unpleasant to you
(16) someone belittling you because of your gender or sexuality
(17) someone excluding you from social gatherings or social network because of your
gender or sexuality
(18) someone asking you for sexual favours in exchange for a reward (e.g. a pay rise
or promotion)
(19) someone threatening punishment or sanctions (e.g. firing) if you refused their
requests for sexual favours
(20) someone touching you against your will (e.g. groped you and/or held on to you)
(21) Attempted rape or actual rape
estimated mean
difference
3.27
0.24
0.59
5.34
4.10
2.83
1.47
1.72
−1.11
−2.14
2.26
2.54
1.53
3.44
−1.54
3.60
4.24
–
–
4.26
–
95% cI low
1.19
−2.10
−2.11
2.49
0.69
−2.71
−3.70
−1.58
−7.15
−7.66
−0.76
−2.23
−1.01
2.12
−7.80
1.56
0.16
95% cI high
5.34
2.58
3.28
8.19
7.51
8.38
6.65
5.03
4.92
3.37
5.29
7.31
4.06
4.77
4.71
5.65
8.32
p
value
0.0020*
0.8394
0.6689
0.0002*
0.0184*
0.3159
0.5766
0.3066
0.7170
0.4457
0.1420
0.2960
0.2370
<.0001*
0.6285
0.0006*
0.0415*
–
–
−1.34
9.85
0.1362
–
All estimates are adjusted for sex and age, and educational level was added as a covariate. results are presented as estimates and 95% cI and *p
<
0.05 as
significant.
cI: confidence interval; IWs: Inventory of Workplace sexual and Gender-based harassment.
harassed, and among those reporting exposure to at
least one item in IWs compared. While coercion is
often considered a more serious form of sexual harass-
ment, research show that frequent expose to gender
harassment is also detrimental to mental health [10].
however, more prospective studies are needed to elicit
the relationship between different types of sexual
harassment and frequency of exposure and mental
health.
IWs builds on a broad understanding of sexual
harassment and includes aspects of unwanted sexual
attention, coercion and gender harassment. IWs also
includes online harassment and harassment related
to employee’s sexuality, which can be seen as a type
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
2843542_0007.png
Two Methods For Assessing Workplace Sexual and Gender-based Harassment
of heterosexism. researchers have more recently
begun to draw parallels between gender harassment
and heterosexism, and rabelo and cortina propose a
closer integration of the two fields [22].
We compared IWs with the self-labelling method
and found that both approaches have advantages and
disadvantages. first, employers may underestimate
the magnitude of the problem and underprioritize
prevention of sexual harassment when relying on the
self-labelling method. An advantage of IWs, and the
behavioural list method, is that it gives workplaces
more detailed data that may focus attention to prob-
lems that might otherwise have been overlooked.
however, we also found some challenges. second,
most of the workplaces in this study were unable to
incorporate IWs in existing employee surveys, for
example, engagement and wellbeing surveys, because
it was too long. Most workplaces chose to use IWs as
a standalone survey, which increases the risk that
workplaces will not continuously follow up, develop a
shorter version or dismiss IWs. Third, small- and
medium-sized workplaces may not be able to ensure
and protect respondents’ anonymity. This problem,
however, will be even more pronounced when using
the self-labelling method.
We collected the data during the covid-19 pan-
demic, with many employees working from home,
which may have resulted in lower prevalence of har-
assment. Although the data were not based on a
representative sample of the Danish population, it
is a strength that we cover different sectors and
industries. The percentage who self-labelled as
being sexually harassed in this study was similar to
previous findings from a representative study
among employees in Denmark from 2018 reporting
a prevalence of 4.6%. The cross-sectional design
does not allow to draw causal inferences whether
sexual and gender harassment has influenced the
level of depressive symptoms, and it possible that
depressive symptoms have caused the reporting of
sexual and gender harassment. finally, the
12-month timeframe may lead to an underestima-
tion between sexual harassment and depressive
symptoms.
conclusion
About 19% of employees reported at least one type
of sexual or gender-based harassment, while 2.5%
self-labelled as being sexually harassed. Both self-
labelling as being sexually harassed and reporting at
least one type of sexual and gender-based harass-
ment in IWs was associated with a higher level of
depressive symptoms compared with being non-
exposed to sexual or gender harassment.
Acknowledgements
7
We thank Anette Borchorst for her valuable input on
the definition of sexual harassment.
Declaration of conflicting interests
The authors declared no potential conflicts of inter-
est with respect to the research, authorship, and/or
publication of this article.
Funding
The authors disclosed receipt of the following finan-
cial support for the research, authorship, and/or pub-
lication of this article: This study was funded by the
Working environment research fund (grant no
2018/42637).
OrciD iDs
https://orcid.org/0000-
Maj Britt Dahl Nielsen
0002-5758-8031
kathrine sørensen
https://orcid.org/0000-0002-
7534-6888
Anna Paldam folker
https://orcid.org/0000-0003-
3952-6918
Per Tybjerg Aldrich
https://orcid.org/0000-0002-
3258-6988
references
[1] Magnusson hanson ll, Nyberg A, Mittendorfer-rutz
e, et al. Work related sexual harassment and risk of sui-
cide and suicide attempts: prospective cohort study.
BMJ.
2020;370:m2984.
[2] McDonald P. Workplace sexual harassment 30 years on: a
review of the literature. 2012;14:1-17.
[3] Nabe-Nielsen k, Grynderup MB, lange T, et al. The role
of poor sleep in the relation between workplace bullying/
unwanted sexual attention and long-term sickness absence.
Int Arch Occup Environ Health
2016;89:967-79.
[4] Nielsen MB and einarsen s. Prospective relationships
between workplace sexual harassment and psychological
distress.
Occup Med
2012;62:226-8.
[5] rugulies r, sorensen k, Aldrich PT, et al. Onset of work-
place sexual harassment and subsequent depressive symp-
toms and incident depressive disorder in the Danish
workforce.
J Affect Disord
2020;277:21-9.
[6] friborg Mk, hansen jv, Aldrich PT, et al. Workplace sex-
ual harassment and depressive symptoms: a cross-sectional
multilevel analysis comparing harassment from clients or
customers to harassment from other employees amongst
7603 Danish employees from 1041 organizations.
BMC
Public Health
2017;17:675.
[7] Blindow kj, Paulin j, hanson lM, et al. sexual and gen-
der harassment and use of psychotropic medication among
swedish workers: a prospective cohort study.
Occup Environl
Med
2022;709:507-13.
[8] carstensen G. sexual harassment reconsidered: the forgot-
ten grey zone.
NORA - Nordic Journal of Feminist and Gender
Research.
2016;24:267-80.
[9] fitzgerald lf, swan s and Magley vj. But was it really
sexual harassement? legal, behavioral, and psychologi-
cal definitions of the workplace victimization of women.
 BEU, Alm.del - 2023-24 - Bilag 127: Orientering om nye forskningsresultater fra NFA om uønsket seksuel opmærksomhed, fra beskæftigelsesministeren
8
[10]
[11]
[12]
M.B.D. Nielsen et al.
In: O’Donohue W, (ed). sexual harassment – Theory,
research, and Treatment. Boston: Allyn and Bacon; 1997.
sojo ve, Wood re and Genat Ae. harmful Workplace
experiences and Women’s Occupational Well-being.
Psychol
Women Q
2015;4:10-40.
cortina lM and Areguin MA. Putting people down and
pushing them out: sexual harassment in the workplace.
2021;8(1):285-309.
fitzgerald lf and cortina lM. sexual harassment in work
organizations: a view from the 21st century.
APA handbook
of the psychology of women: Perspectives on Women’s Private and
Public Lives, Vol
2.
APA handbooks in psychology®.
Washing-
ton, Dc, us: American Psychological Association; 2018.
pp. 215-34.
Ilies r, hauserman N, schwochau s, et al. reported inci-
dence rates of work-related sexual harassment in the united
states: using meta-analysis to explain reported rate dispari-
ties.
Pers Psychol
2003;56:607-31.
Magley vj and shupe eI. self-labeling sexual harassment.
Sex Roles
2005;53(3):173-89.
The labour Market Inspectorate D [Internet] 2022. Avail-
able from: https://at.dk/arbejdsmiljoe-i-tal/arbejdsmiljoe-og-
helbred-2012-2018/
fitzgerald lf, Michele Gf and Drasgow f. Measuring
sexual harassment: theoretical and psychometric advances.
Basic App Soc Psych.
1995;17:425-45.
Gutek BA, Murphy rO and Douma B. A review and cri-
tique of the sexual experiences Questionnaire (seQ).
Law
Hum Behav
2004;28:457-82.
latcheva r. sexual harassment in the european union: a
pervasive but still hidden form of gender-based violence.
J Interpers Violence
2017;32:1821-52.
cuenca-Piqueras c, fernández-Prados js and González-
Moreno Mj. face-to-face versus online harassment of euro-
pean women: importance of date and place of birth.
Sex Cult
2020;24:157-73.
Nielsen MB, Bjørkelo B, Notelaers G, et al. sexual harass-
ment: prevalence, outcomes, and gender differences assessed
[21]
[22]
[23]
[24]
by three different estimation methods.
J Aggress Maltreat
Trauma
2010;19: 252-74.
Berdahl jl. harassment based on sex: protecting social
status in the context of gender hierarchy.
Acad Manage Rev
2007;32:641-58.
rabelo vc and cortina lM. Two sides of the same coin:
gender harassment and heterosexist harassment in lGBQ
work lives.
Law Hum Behav
2014;38:378-91.
silverschanz P, cortina lM, konik j, et al. slurs, snubs, and
queer jokes: incidence and impact of heterosexist harass-
ment in academia.
Sex Roles
2008;58:179-91.
Bech P, rasmussen NA, Olsen lr, et al. The sensitivity and
specificity of the Major Depression Inventory, using the
Present state examination as the index of diagnostic valid-
ity.
J Affect Disord
2001;66:159-64.
Olsen lr, jensen Dv, Noerholm v, et al. The internal
and external validity of the Major Depression Inventory
in measuring severity of depressive states.
Psychol Med
2003;33:351-6.
Bech P, Timmerby N, Martiny k, et al. Psychometric evalu-
ation of the Major Depression Inventory (MDI) as depres-
sion severity scale using the leAD (longitudinal expert
Assessment of all Data) as index of validity.
BMC Psychiatry
2015;15:190.
Olsen lr, Mortensen el and Bech P. Prevalence of major
depression and stress indicators in the Danish general popu-
lation.
Acta Psychiatr Scand
2004;109:96-103.
øhrstrøm Be j and knudsen l. undersøgelse af forekomst
og oplevelse af kønskrænkende adfærd i forsvaret. hovedresul-
tater [Investigation of the prevalence and experience of gender
harassment in the army]. forsvarsakademiet 2003.
stark s, chernyshenko Os, lancaster Ar, et al. Toward
standardized measurement of sexual harassment: shorten-
ing the seQ-DoD using item response theory.
Mil Psychol
2002;14:49-72.
Murdoch M and McGovern PG. Measuring sexual harass-
ment: development and validation of the sexual harassment
Inventory.
Violence Vict
1998;13:203-16.
[13]
[25]
[14]
[15]
[16]
[17]
[18]
[19]
[26]
[27]
[28]
[29]
[30]
[20]