A Cross-Sectional Study of Educational Aspects and Self-Reported Learning Difﬁculties among Female Prisoners in Norway

: The aim of this cross-sectional study was to analyse the educational background, educational desires and participation in education among three samples of female prisoners with Norwegian citizenship in Norwegian prisons over the period from 2009 to 2015. The female participants were n = 106 in 2009, n = 74 in 2012 and n = 79 in 2015, respectively, with a mean age of 38 years. Moreover, the study examined whether self-reported learning difﬁculties could predict participation in education activity while incarcerated. The results show that the female prisoners included in this study increased their educational level over the studied years. Similar education patterns were observed in the 2009 and 2012 samples regarding all educational levels for the female prisoners. A different pattern was observed in the 2015 data, with 44.3 % having mandatory education as their highest level compared to 57.6 in 2009 and 53.4 in 2012, respectively. However, these differences in percent between the samples at any education level were not signiﬁcant. Signiﬁcant differences were, however, found regarding the desire for upper secondary education between the samples in 2009 and 2012, and 2009 and 2015, respectively. In 2009, 20.2% reported upper secondary education as an educational desire, whereas 35.2% reported this as a desire in 2012, and 36.7% in 2015. Participation in educational activity during incarceration also changed during the time period of these studies. Many of the female prisoners participate in educational activity, but a signiﬁcant difference was found between the samples in 2012 and 2015 as there was a decrease in activity. In 2012, 41.9% did not participate, whereas in 2015, almost 60% (58.2) of the female prisoners did not participate in any educational activity. Both the highest completed education level and self-reported learning difﬁculties predicted participation in education activity among the female prisoners in the 2015 sample.


Introduction
There are several reasons for providing education to prisoners. Education for all, including incarcerated individuals, is an important facet of civilised society, and it contributes to the development of the whole person. There is also an increasing awareness of the importance of educating prisoners to improve their life quality and reduce recidivism [1,2]. Several studies characteristically show a reduced likelihood of recidivism among prisoners who participate in educational programmes compared to prisoners who do not participate [2][3][4][5]. Research has revealed that in addition to benefits to the individual prisoner, the social return to schooling is in fact larger than the private return [6].
Quality assured data on female prisoners are less available than research on prisoners in general. In fact, there is a lack of empirical studies including factors such as female prisoners' education level, their educational desires or actual participation in education while incarcerated. Moreover, we also lack information on factors potentially influencing participation in education among female prisoners, such as learning difficulties (reading, writing and mathematical). To our knowledge, a cross-sectional study of educational issues among female prisoners has not yet been reported. This paper aimed to place a special focus on female prisoners and their characteristics, both internationally and nationally. The main aim was to analyse the female prisoners in Norway and their educational background, educational desires, learning difficulties and participation in prison education over the period from 2009 to 2015. The data are discussed with respect to female prisoners' needs and requirements.
Globally, the percentage of female prisoners is 6.9 [7]. The number in Norway is equivalent, as the female population makes up 6% of the prison population in Norway; however, the incarceration rates are higher, as 9.3% of the prisoners are females, indicating that female prisoners serve shorter sentences [8]. Due to the low rates of females in prisons, there has historically been little consideration of their specific needs [9], and prison systems are typically organised around the needs and requirements of male prisoners [10,11]. According to Coyle [11], this applies to architecture, security and all other facilities: "There is a recurring tendency that any special provision for women prisoners will be something that is added on to the standard provision for men. This is despite the fact that the profile of women prisoners is very different from the male prisoners, and particular attention should be given to their special needs" ( [11] p. 143). Bartels and Gaffney [10] point out that it is important to consider female prisoners' pathology when building education models and programmes specifically for women to enhance their self-esteem and general functioning. For female prisoners to achieve their rehabilitative goals, they need prison models that provide support, services and requirements especially for women, as this can help reduce recidivism and support their daily life functioning, during and after incarceration. They summarise their literature review on female prisoners into two main points: one concerns the need to recognise the gender-specific needs of female prisoners and the other that the rehabilitation of female prisoners must be gender-specific and not just as a complementary model suited for male prisoners [10].
Research among prisoners has often been conducted using the whole prison population as one group, where the male prisoners usually comprise 90% of the participants. Binswanger and colleagues' [12] study focusing on gender differences in the prison population showed that male and female prisoners differ in central demographic variables, such as age, race, civil status, education, work and living conditions. Ellis, McFadden and Colaric [13] also emphasise the importance of the unique characteristics of prisoners, such as gender, strengths and special challenges. Such empirical data could provide necessary information to improve interventions in prisons in the areas of prisoner background regarding needs and employability after release, as well as best practices to improve the success of correctional educational programmes for females [13].
Several studies from Australia show that the female prison population has changed. There are more female prisoners with mental health problems, drug-related problems and social deprivation compared to the females in the general population [14][15][16]. According to Stathopoulos et al. [14], female prisoners are a marginalized group with a high percentage of metal health problems, sexual abuse, both in childhood and adulthood. They have a low socioeconomic status, low cognitive skills, a low education level and supporter responsibility for children. Compared to male prisoners, they report that female prisoners have minimal work experience. These characteristics of female prisoners are, according to Stathopoulos et al. [14], independent of whether the studies are British, Scottish, American or Australian. These findings are also supported by Fair [17], who in her review article revealed similarities between female prisoners across nations and outlined that female prisoners are among the most socially deprived groups in society. In a study of prisoners in Sweden, Lindberg [18] reports that female prisoners had more social and economic challenges compared to male prisoners.
The percentage of female prisoners in Norway has remained stable, between approximately five to just over six percent (6.2%) [19]. Female prisoners in Norway are placed in a) prisons exclusively for females (four prisons) or mixed gender prisons (five prisons) [20]. Female prisoners have separate needs to the male prisoners, and it is important that the prisons facilitate female-specific needs, which implies both physical conditions (buildings) and contentwise facilitations [21,22]. Several national reports have, in recent years, pointed out that female prisoners do not have the same offers when it comes to physical activity and outside areas and are not offered the same quality of addiction therapy as male prisoners. Moreover, some of the reports point out that there a very few places with low security for females. This can challenge the correct use of security levels [23][24][25][26][27]. According to Vige's report [26], female prisoners in mixed prisons had been encouraged to participate in education to a lesser extent compared to the female prisoners in female prisons. Furthermore, those with a low education level, elderly prisoners, and those with residence outside of Norway were also to a lesser extent encouraged to participate in education while incarcerated.
Studies on the prison population in Norway report a high percentage of mental problems, with female prisoners reporting the highest rates [28,29]. Moreover, in Amundsen's [30] study, where 100 females were included, four out of ten reported to have been abused as a child, eight out of ten reported serious depression and six out of ten reported having drug-related problems, which resulted in great economic, social and health challenges. Based on her findings, Amundsen [30] indicated that the female prisoners are a socially marginalised group. Furthermore, in her comparative study of male and female prisoners, Amundsen [31] reported that there are significant gender differences as female prisoners self-report more mental health (anxiety) problems before sentencing. Cramer [32] examined a representative sample of prisoners' mental health problems (N = 857) and found no significant gender differences within the main groups of symptom disorders, such as personality disorder or ADHD. However, Cramer [32] found considerable differences in the reporting of mental health problems between the prison population and the general population.
The life-long learning perspective, which also includes education for prisoners, is emphasised by the European Common Frameworks' eight competencies for lifelong learning [33]. According to international conventions and recommendations, education is recognised as a basic human need and human right, including for prisoners, and prisoners with learning difficulties should be given particular attention [34]. Several key documents are relevant for the situation of females who are incarcerated, among them the Standard Minimum Rules (SMR) for the treatment of prisoners [35] and the European Prison Rules from 1973, revised in 2006. Rule 34.1 in the European Prison Rules states that in addition to the concrete provisions for female prisoners, the state shall place special focus on the female prisoners' physical, work-related and psychological needs when it comes to decisions concerning a female's incarceration situation [24,36]. A revised version of SMR in 2015, The Mandela Rules, includes its own recommendation for female prisoners [37]. Although these guidelines are not legally binding, they outline important recommendations for the female prisoners and can contribute to an equal correctional service.
Norway and several countries (e.g., the Nordic countries Denmark, Finland, Iceland and Sweden) have incorporated the European Convention for the Protection of Human Rights and Fundamental Freedom [38] into their legislation, and both primary and secondary education is offered in prisons. This implies 7 years of mandatory primary school (age 6-13), 3 years of mandatory lower secondary school (age 13-16) and 3-5 years of upper secondary school (age [16][17][18][19]. Upper secondary education is a legal right, but it is not mandatory. The Execution of Sentences Act [39] states that the individuals in custody should be assured adequate civil services through collaboration with providers in the community. In addition, education is included as part of the obligatory participation in activities during the custody ( §3) [39]. Access to social services and education should be equivalent to other citizens. The provision in Norway is organised by using the so-called "import model" [40]. The school in the local community also provides the educational programmes in the prison. For example, the prisoners receive their diploma from the local school. The Education Act [41] gives the prisoner in primary and secondary education the right to an education in line with their needs ( §1.3). Moreover, they have a right to special education if they cannot receive a satisfactory outcome from ordinary education ( §5.1). In Norway, the prisons have currently established educational programmes at the mandatory level and upper secondary level, and at these levels, there are formally qualified teachers. In addition, prisoners have access to higher education (single subjects or degrees) through distance learning, although some prisoners in low-security custody can also be offered day release to participate in courses.
This is reason to believe that having various learning difficulties is part of the underlying cause of problems in school for some prisoners. In this respect, examining learning difficulties among prisoners may add to our understanding and provide knowledge that may be used in the development of a better support system for those who have such difficulties. Previous studies from the entire Norwegian prison population have revealed that less educated prisoners and prisoners with learning difficulties may see their time in prison as a "second chance", an opportunity to complete an education. In one study, having literacy difficulties more often predicted the desire of prisoners to participate in upper secondary education compared to those without such difficulties, when other factors were controlled for [42]. Additionally, studies have reported that prisoners who self-report a diagnosis of dyslexia are more likely to participate in prison education [43]. Moreover, a prisoner's actual level of reading and writing skills, measured by objective tests, did not predict participation in education [43]. Thus, these mentioned studies did not include gender-specific analysis, so the issue of whether learning problems predict participation in prison education remains to be addressed for the female prison population in Norway.
The main aim of this cross-sectional study was to address female prisoners' educational pattern over a period of seven years through three data points by reporting their educational background (highest completed level), educational desires and educational activity during incarceration. Second, the study aimed to analyse whether female prisoners' self-reported learning difficulties predict participation in education during incarceration. More specifically, the research questions are as follows: (1) Was there a change in educational pattern between 2009 and 2015 for female prisoners in Norway? (2) Can learning difficulties predict educational activity while incarcerated?

Materials and Methods
This study was part of a larger study on prisoners' educational competence in all three time periods. Inclusion criteria for taking part in the present study were females over 18 years old with Norwegian citizenship. The study included three samples of female prisoners from the years 2009, 2012 and 2015. In the three samples, women accounted for 7.7, 5.8 and 5.0%, respectively, of the prisoners with Norwegian citizenship included in the study (n = 106 in 2009, n = 74 in 2012 and n = 79 in 2015, respectively). Mean age was 37.8 (sd = 11.6), 37.5 (sd = 11.0) and 38.1 (sd = 11.1), respectively, for the females in 2009, 2012 and 2015. The overall response rate was 63.2% in 2009, 52.3% in 2012 and 56.3% in 2015, respectively.

The Questionnaire
A questionnaire was sent out to all prisons in Norway one specific week in October in 2009, 2012 and 2015. Questions in the questionnaire relevant for the present study included demographic variables such as age, sentence length, highest level of completed education, learning difficulties, educational desires and educational activity in prison. The questions on self-reported learning difficulties were measured on 4-point scales, where the participants rated their reading, writing and mathematics difficulties (on separate scales) using the graded scales: "yes, large difficulties", "yes, difficulties to some extent", "yes, some difficulties" and "no difficulties". For most of the questions, the respondents were asked to tick the appropriate box or boxes. The three variables on self-reported learning difficulties, reading, writing and mathematics were made into one variable ("learning difficulties") in the logistic analyses.

Procedure
A person representing the County Governor (Province Governor) of Vestland and the Department of Education (the organisation in charge of Norwegian prison education, serving the Ministry of Education) telephoned each prison governor and each headmaster in charge of prison education in order to outline the purpose of the study and to arrange for the assessment to be carried out. In addition, a letter was sent to the same persons, explaining the procedures. In line with instructions from the research group, the prison governor of each prison or the teacher in charge of education carried out the survey. Prisoners with reading and writing difficulties were offered assistance by prison teachers or prison officers in completing the questionnaire.

Ethical Considerations
The studies that this study bases its data on were all approved by the Data Protection Official for Research, NSD (Norwegian Centre for Research Data), and additional approval was granted by the prison authorities and the Ministry of Justice and Public Security. The front page of the questionnaire explained the purpose and procedure and emphasized that participation was voluntary. By completing the questionnaire, the prisoners consented to participating in the study. In the questionnaire, it was explicitly stated that the information provided would be confidential and that individual information would not be used by the prison authorities. No compensation was given for participation as this could have put them under pressure to reply. It was made explicit on the front page of the questionnaire that the respondents could withdraw from the study at any time without any consequences.

Statistical Analyses
Descriptive data from the data sets in 2009, 2012 and 2015 are presented for the females' educational background, desires and participation in education while incarcerated. Mean age is reported from the three samples. Correlations among (Pearson product moment) age, educational level (highest completed), participation in educational activity and selfreported learning difficulties are reported for the samples in 2012 and 2015 in Table 1. To address whether self-reported learning difficulties could predict participation in prison education, logistic regression analyses were performed from the two samples in 2012 and 2015. Due to some differences in variables in the 2009 data, this data set was not included in all analyses. All statistical analyses were conducted using SPSS version 25. Note: Education level = highest completed education level. Self-reported difficulties = reading, writing and mathematics. Educational activity = participation in educational activity. ** Significant at p < 0.01, * significant at p < 0.05.

Results
The three samples were similar in age range. Mean age was 37.8 (sd = 11.6), 37.5 (sd = 11.0) and 38.1 (sd = 11.1) in 2009, 2012 and 2015, respectively. Table 1 provides an overview of the correlations among age, educational level (highest completed) self-reported learning difficulties and education activity for the female prison samples in 2012 and 2015.
In the 2012 sample, there was only a significant negative moderate correlation between education level and self-reported learning difficulties. In the 2015 sample, a significant positive correlation was revealed between age and education level, and significant negative correlations were observed between education activity and highest completed education level and between age and education activity. Tables 2-4 show an overview of the percentages of female prisoners' highest level of completed education, educational desires and their participation in educational activity during incarceration. Differences between the samples were calculated using MedCal (MedCalc's Comparison of proportions calculator). Significant differences are reported under each Table and in the text.     (Table 2). However, there were no significant differences between the samples at any education level.
Significant differences were found regarding the desire for upper secondary education between the samples in 2009 and 2012 and 2009 and 2015, respectively. One in five (20.2%) reported upper secondary education as an educational desire in 2009, whereas more than one in three had this desire in 2012 (35.2%) and 2015 (36.7%). Participation in educational activity during incarceration also changed over the time cycle of these studies. A significant difference was found between the samples in 2012 and 2015. In 2012, 4 out of 10 did not participate, whereas in 2015, 6 out of 10 did not participate in any education activity (Table 4). Moreover, a significant difference was found between the samples in 2012 and 2015 regarding "other education". As observed in Table 4, the percentage of prisoners attending other courses decreased from 25.7 to 12.7%.
To explore whether self-reported learning difficulties (including all difficulties: reading, writing and mathematics) had an impact on participation in education while incarcerated, logistic regression analyses were performed (Table 5). In both analyses, the dependent variable was participation in education (coded 1) and no participation (coded 0). The predictor variables (independent) were age, self-reported learning difficulties and highest completed level of education. In the 2012 data, the analysis revealed that none of the predictor variables were significantly related to whether the females participated or not in education. In contrast, in the 2015 data, analysis showed that self-reported difficulties had a significant (p < 0.02) positive effect on participation in education (Exp(B) = (1.97). The higher the number of self-reported difficulties, the more likely it was that the prisoners participated. Age also had a significant (p < 0.05) effect, as younger prisoners were more likely to participate (Exp(B) = (0.46). However, highest completed education level was not a significant predictor for participating in education in the 2015 data.

Discussion
The aim of the present study was to explore possible changes in educational patterns between 2009 and 2015 for three samples of female prisoners with Norwegian citizenship. Moreover, the study examined whether self-reported learning difficulties can predict educational activity while incarcerated for the two samples of female prisoners in 2012 and 2015. A moderate negative correlation between the highest completed education level and self-reported difficulties among the 2012 sample reveals that female prisoners with the lowest education level self-report more learning difficulties; this relation was, however, not observed in the 2015 sample. These results suggest that the type of educational difficulty female prisoners report has some level of variability between cohorts over time and that reported persistent ongoing leaning difficulties are a different concern to the lack of opportunity to learn. Moreover, the results also suggest that self-measures may only be one source of information in determining the extent of educational difficulties among prisoners. Additionally, the small samples in the study and the variation within the three samples can influence the significance level of the analyses. In 2015, age correlated positively with education level, and negative correlations were observed between the highest completed education level and education activity, and between education activity and age. There was a significant difference between the prisoners' desire for upper secondary education from 2009 to 2012/2015. Participation in educational activity showed a significant difference between the sample in 2012 where 4 out of 10 did not participate, and the 2015 sample, where almost 6 out of 10 did not participate in an educational activity. Moreover, a significant difference between the 2012 sample and 2015 sample in participating in "othereducation" was also observed. Prisoners who self-reported learning difficulties and younger prisoners were more likely to participate in education in the 2015 sample, whereas in 2012, these variables did not predict participation. There is a general decrease in educational activity from 2012 to 2015 and specifically the category "other education". This may be due to several reasons. A natural explanation can be that the educational offers in prison education cannot meet their needs for, e.g., higher education. Eikeland et al. reported that prisoners participating in upper secondary education were more satisfied than those participating at university or college levels.
One third of the prisoners wanted to take upper secondary education in 2015, and only one in four participated at this level. Although prisoners have the right to education (The Educational Act), they may, due to their short sentence, not find it worth embarking on. The mean sentence length among prisoners in Norway is 170 days [20]. A short sentence may be a hindrance for participating in education, and previous studies have found that sentence length had a significant negative impact on participating in education while incarcerated [43,44]. The lack of available programme options for female prisoners may be another important factor at the system level. According to Vige [26], female prisoners in Norway were encouraged to a varying degree to participate in education, especially those in mixed prisons, those with a low educational level, elderly prisoners and those with residence outside of Norway. Moreover, several national reports from Norway have outlined the need for gender-specific adaptations, and some of the reports point out that there are very few places with low security for females and that this challenges the right use of security levels (cf. [24,25]). When female prisoners are given a higher security level, this may also affect their educational options. However, figures from the correctional database indicate that females in custody have participation rates similar to male prisoners in most activities, but shorter sentences and higher education may make some of the programmes offered irrelevant [45]. Many barriers to education experiences by female prisoner, such as access to computers and internet, lack of information and lack of access to relevant programmes, seems to be a general challenge in Norwegian prisons [46].
In Rose's study [4], among female prisoners in the US, several factors are outlined that can affect women's participation in educational programmes and explain why some do not participate, even though it has a positive effect on recidivism, motivation and pro-social behaviour. The lack of available programme options for female prisoners is outlined by Rose [4] as one important factor for not participating in education. Moreover, he found that women who received visits from their children were significantly more likely to participate in education programmes compared to those women who did not. Additionally, those who participated in educational programming were more likely to participate in other types of programmes offered in the prison [4]. Due to the few female prisons in Norway, female prisoners are often incarcerated far away from their families. This can result in fewer visits from children and family members [25]. Rose [4] found that incarcerated mothers' separation from their children is one of the worst aspects of imprisonment (cf. [47]), and the separation can affect their mental health and their coping with education. Previous studies show that female prisoners report a high amount of mental health problems [28,29,31]. This can result in female prisoners not participating in education programmes. Mental health problems, drug-related problems and, for example, sleep disorders might, therefore, explain the gap between the number of female prisoners who want to participate in education and the number who participate. The abovementioned studies found that female prisoners report more mental health problems than male prisoners, but in Cramer's study [32], there were no significant gender differences in reporting mental health problems.
Granger-Brown [9] investigated factors that invite female prisoners to engage in learning opportunities and tools and techniques that may support or enhance their learning process. The author found that women are motivated to learn and want educational opportunities, and they believe that they can learn and develop if given access to information. The results from the study also showed that feeling emotionally safe in a positive environment is an important factor in the learning process. Many of the women interviewed in the study stated that the learning process and the incarceration were important for them to change their lifestyle and take control over their life [9]. Although the percentage of prisoners desiring "other education" (this includes short courses) was stable over the seven years in which we conducted our studies, there is an increasing gap between those desiring "other education" and those reporting having participated in this educational alternative. In the 2012 sample, the relation between desiring and participating in "other education" was almost the same. In contrast, in the 2015 sample, one in four desired "other education", but only half of them (12.7%) were participating in that option.
Self-reported learning difficulties and age were both significant predictors for participating in education among the 2015 sample, but not in the 2012 sample. At the individual level, self-reported difficulties, as a predictor for participating in education, may be explained by the fact that prisoners see their time in prison as a "second chance" to a better adapted education compared to the ordinary school. Moreover, in prison, the classes are smaller, and they may also find more peers at the same level as themselves. These findings and interpretations are in line with previous studies on the prison population in Norway [43] and social comparison theory [48]. Studies have also reported a relationship between age and self-reported difficulties as younger prisoners report more reading and writing difficulties than older prisoners [49]. As self-reported difficulties and age only predicted participation in education in the 2015 sample and not in the 2012 sample, the results may have system-level explanations. Is there a lack of resources so only prisoners who are characterised as prioritised groups gain access to education? According to international recommendation, prisoners with learning difficulties should be given particular attention [34]. Moreover, younger prisoners are also a group who are explicitly prioritised for education [36]; see rule 28.3 and §18 of The Execution of Sentences Act [39]. Furthermore, there are concerns regarding the changes in prison population; the prison population is more demanding and serves longer sentences, as many of those with short sentences are executing a sentence with electronic monitoring [50]. Although the abovementioned changes are not specific to the female prisoners only, the changes may affect them more as they represent a minority in prisons. Prison systems are, according to Hawley and colleagues [51], geared towards male prisoners and, therefore, tend not to take into account the different situations and problems of women.
In sum, based on the data in the present study, we find that the educational level increased at all levels, but there was also a decrease in general educational activity over these years. There is also an increasing gap between desiring an education and actual participation in upper secondary education and the category "other education". These tendencies are also observed among male prisoners (e.g., [49]).

Limitations and Implications of the Current Study
Some limitations of this study need to be addressed. As observed from the three samples, female prisoners constitute a small number of the total prison population in Norway. Nevertheless, the female prison population in Norway, like in most other countries [7], was stable over the period in which the data for the present study were collected. With sample sizes between n = 73 and 106, calculated percentages include a small number of prisoners. The significant correlations presented in Table 1 are moderate. Although the study includes the whole population of female prisoners with Norwegian citizenship, there is a risk for type 2 errors when the sample size is small. Notwithstanding the abovementioned limitations, the results of the present study provide useful insights into female prisoners' education over a period of seven years. When discussing and comparing studies on female prisoners across nations, one must bear in mind the differences in sentencing conditions, the diversity between prison populations and within a prison population in one nation and the educational offers in prisons. Previous studies on education and learning difficulties in Norway have reported that female prisoners resemble male prisoners with regard to educational level and self-reported learning difficulties [49,52].
To sum up findings from international studies with national research from Norway, we found that the characteristics of female prisoners and their needs, as outlined in the studies and reports, share similarities across nations. Thus, the picture is not unambiguous, as sample sizes and sample selection may affect the results. Addressing features of female prisoners in research is important, and gender-specific research among prisoners can be a way of improving prison education so that it is best suited to incarcerated women's needs. Being gender responsive in correctional services requires us to acknowledge the realities of women's lives, which includes knowledge about their criminal pathways and the relationships that shape their lives [53]. A knowledge-based correctional service will be important to meet the diverse needs of prisoners. Prison education can be a valuable new start for prisoners, and empirical findings on educational variables and learning difficulties can provide valuable information about female prisoners' educational patterns and changes. We encourage further studies on female prisoners as their own group with representative samples. The findings in the present study also point out the importance of following educational trends over time in a prison population. A new study on female prisoners in Norway and Belgium covering education, motivation and learning challenges will be completed in the spring of 2021. This study has a mixed method design with survey data; individual testing of reading, writing and language skills; and focus group interviews. Further investigation should allow for comparative studies between two female populations but also further gender comparisons between similar data from male prisoners in Norway.