Individual educational plans: Just a tool to immunise teaching from parental criticism?

Abstract Categorised as planning documents, individual educational plans (IEPs) purport to provide school-based instruction more closely matched to individual pupils’ skills and needs. The present study investigated the extent to which IEPs contain practically relevant suggestions for teachers to implement in practice. Based on a sample of 112 IEPs from Germany, the support proposals in these documents were examined. Results showed that about half of the entries consisted of conventional, content-specific methods to be carried out in school. In the average IEP, only 30 per cent of the entries contained content that deviated from the curricula for a particular grade level. Another large proportion of support proposals consisted of conventional but content-unspecific methods to be carried out in school. These proposals were mainly catchwords of generic differentiated instruction practices, signalling adequately personalised support towards to the IEPs’ readers but lacking sufficiently clear descriptions to ensure their implementation. Analyses of variance concerning specific contextual and structural characteristics of IEPs made it possible to assess the features of documents with higher potential informational value for teachers and support staff. Overall, most IEPs seem unlikely to fulfil their primary purpose. At the same time, there were indications about how they could be improved.


Introduction
A good individual educational plan (IEP) is a "written curriculum-based pedagogical document, intended to support a student's learning process and growth" (Räty et al., 2019, p. 25). Creating IEPs is meant to be useful for teaching pupils with disabilities, preventing learning and behavioural problems, and supporting gifted pupils (Ninkov, 2020;Popp et al., 2017;Xu & Kuti, 2021). Whether it is compulsory or just recommended to compile an IEP in such situations depends on legal requirements that differ across countries and states (Florian, 2021;Mitchell et al., 2010;Nastiti & Azizah, 2019). Apart from personal data on the pupil and, frequently, medical diagnoses as well as controversial school administrative labels, IEPs usually contain four core elements that build upon each other: a) a description of the present level of performance, b) a definition of intended learning goals, c) recommended supportive actions, and d) conditions and methods for the evaluation of goal achievement (Özdemir et al., 2020, p. 130;Al-Shammari & Hornby, 2020, pp. 168-169;Popp et al., 2017, pp. 30-31;Andreasson et al., 2013, p. 413;Sanches-Ferreira et al., 2013, p. 507;Ruble et al., 2010Ruble et al., , p. 1463Hauer & Feyerer, 2006, p. 122;Xu & Kuti, 2021, p. 2). This logical structure sounds plausible and straightforward. Hence, one can easily imagine putting in individual descriptions of the learning level, goals and support proposals that most pupils, even younger ones, could understand, discuss and act upon (Blackwell & Rossetti, 2014;Goepel, 2009). Nevertheless, teachers tend to consider IEPs and similar documents to be of little use compared to the effort required to create them Hirsh, 2014;Moser Opitz et al., 2019;Müller et al., 2017). What systematic reasons could undermine them in attaining their intended purpose of being "effective and functional and used by teachers and support staff to inform their daily practices" (Muñoz Martínez & Porter, 2018, p. 4)? Furthermore, IEPs should be the basis for mutual exchange and coordination, especially between parents and teachers: How, then, could the findings be explained that parents "are often not provided the opportunity to make significant contributions to the content of their children's IEPs" (Blackwell & Rossetti, 2014, p. 9)?
Clarifying these questions seems practically urgent. Well-written IEPs would make an essential contribution in ensuring inclusive education (e. g. King et al., 2018;Räty et al., 2019). They should be used to initiate adaptations from standard pedagogical support and, thus, overcome barriers (Ainscow, 2021). The present study attempts to assess these questions empirically by examining to what extent IEPs are informative for everyday school practice. The results indicate how IEPs can be made more informative and include recommendations on what questionable content can and should be omitted to avoid degradation of pupils and consternation on the part of parents.

How does the IEPs' content relate to teaching the "rest" of the class?
From the four above-mentioned document-typical contents, it is support proposals in IEPs that may have an exceptionally high practical relevance for everyday practice. It is the rubric where school staff, pupils, and parents should find individualised instructions for learning and social participation support. Entries under this rubric might refer, at least indirectly, to the background of teaching the entire class because "[a]n IEP should record what is 'additional to' and 'different from' the teacher's regular differentiation planning" (Goepel, 2009, p. 126). Hence, it is surprising that mentions of support proposals are far from being present in every IEP, as Rakap et al. (2019) found in a sample of 206 IEPs from Turkey. Similar findings based on a smaller, US-American sample were observed by Ruble et al. (2010Ruble et al. ( , p. 1464). An investigation based on a selection of IEPs from Turkey indicates that these documents "did not provide a sufficient level of information to organise the instruction" (Özdemir et al., 2020, pp. 133-134). Räty et al. (2019) explicitly explored the "support measure utterances" (referred to as support proposals in this text) in 21 IEPs from Finnish pupils with intellectual disabilities. Their study defined support measures as "all pedagogical methods aimed at supporting students' learning, such as arrangements for the learning environment, teaching and working methods, aids, modifications and accommodations" (Räty et al., 2019, p. 39). They found that only a few proposals referred to "Learning objectives", while the clear majority could be subsumed under the categories "Teaching adjustments", "Individual guidance", and "Learning environment" (Räty et al., 2019, p. 41). Most of the proposed methods correspond to what can be described as differentiated instruction (DI; Pozas & Schneider, 2019). Particularly noteworthy is the finding that although the IEPs were for pupils with intellectual disabilities, most support proposals followed "general principles" of teaching (Räty et al., 2019, p. 41). Above all, Räty et al. stated "that most support measures were not described thoroughly" (Räty et al., 2019, p. 44).
In many regions, special needs teachers are present only a few hours per week in an inclusive classroom (Faldet & Nes, 2021;Klemm, 2020;Nilsen, 2017;Steinmetz et al., 2021). For Finland, Nilsen found evidence that regular and special educational needs teachers typically work separately (Nilsen, 2017). Instead of adaptations in terms of content, work plans for pupils with special needs primarily provided simpler adaptations, e. g. concerning the number of tasks (Nilsen, 2017, p. 210). Hauer and Feyerer explored 50 IEPs from Austria. None of these documents established a relationship between the whole class and the recommended supportive actions for the child the IEP was compiled for (Hauer & Feyerer, 2006, p. 139). In their investigation of goals in the IEPs of 125 Swiss pupils in inclusive classrooms, Paccaud and Luder found that "[m]ost goals concerned academic topics (56 %), and two thirds of the children had goals corresponding to the curricular level" (Paccaud & Luder, 2017, p. 205). Besides the academic goals, they classified a broad set of "functional goals", such as "organizing routines" and "applying the knowledge that is learned" (Paccaud & Luder, 2017, p. 213). The large portion of academic goals and their percentage of grade-level accordant ones could be interpreted as high accessibility of regular teaching to the pupils with IEPs. At the same time, it remains an open question to what extent such goal formulations as well as the descriptions of many of the "functional goals" are, at least from teachers' perspective, redundant.

Teachers' mission and the constraints upon resources as an implicit perspective in IEPs
Teachers are responsible for considering "what elements of the lesson can be the same and what specific adjustment or alternative instructional strategies may be needed" (Muñoz Martínez & Porter, 2018, p. 10). In order not to make lessons unmanageably complicated for themselves, teachers could simply choose not "to plan and deal with academic goals which are under the level of the curriculum; therefore are tempted to set fuzzy goals or prefer to plan functional goals" (Paccaud & Luder, 2017, p. 221). By committing only to relatively few apparent deviations from the whole class's teaching in the IEP, there might be less pressure to prepare every lesson for several learning levels or with other profound adaptations (Fields, 1999, p. 12;Greiten, 2017). Simultaneously, there already is a common demand to make content accessible to the entire class in its heterogeneity of individual pupils (Everitt, 2017). For this, suitable individualised adaptations and differentiated instruction (DI) methods, e. g., "tiered assignments" and other widely known practices (Pozas & Schneider, 2019), should be employed. Moreover, knowledge of these methods is not exclusive to special needs teachers. Recent studies indicate that regular subject teachers also use DI methods, if not always, then at least regularly (Labhart et al., 2018;Letzel & Otto, 2019;Pozas et al., 2020). Adaptive and differentiated instruction are considered desirable and effective (Deunk et al., 2018;Gheyssens et al., 2020;Smale-Jacobse et al., 2019).
However, the crux of the matter is concretising these initially abstract guidelines. Adaptive teaching methods are not fundamentally different from standard teaching methods (Parsons et al., 2018, p. 229). Differentiated instruction methods are also generic in themselves and must be filled with content for individual pupils or subgroups of the class in concrete situations. There are numerous indications that teachers do not feel sufficiently trained for this and that the interventions on differentiated instruction that have been evaluated as effective were instead designed by the research teams of the respective studies (Deunk et al., 2018;Smale-Jacobse et al., 2019). Therefore, raising awareness and adapting instruction to pupils' different needs is undeniably important. At the same time, it is necessary to consider the complexity of this task (Van Geel et al., 2019), the largest part of which is to be tackled by teachers.
Frequently, IEPs are found to be deficit-oriented (Andreasson & Carlsson, 2013;Isaksson et al., 2007). In these cases, they implicitly indicate a degree of the school's incapability to address and sufficiently include particular pupils adequately. If this is understood, at least partially, as a projection, it fits with parents' perception of IEP meetings as burdensome (Scanlon et al., 2018;Zeitlin & Curcic, 2014). It is widely known that these documents are intended to serve at the same time "educational, legal, planning, accountability, placement and resource allocation purposes" (Mitchell et al., 2010, p. iii; see also, Boban & Hinz, 2007;Moser Opitz et al., 2019). Thus, IEPs also may be a tool by which the school as an institution displays its efforts to support particular pupils (Andreasson & Carlsson, 2013;Isaksson et al., 2007).

A way to understand IEPs: Assessing their informational value
Teaching the whole class is the task into which most planned and individualised supportive actions must be integrated (Finkelstein et al., 2021;Nilholm, 2020;Xu & Kuti, 2021). However, the meaning of this prerequisite has not yet been sufficiently taken into account when analysing authentic IEPs. Whether an entry in an IEP may be perceived as informative or not from teachers' perspective can be estimated by a specified application of Bateson's formal definition of "information". He describes "information" as "any difference which makes a difference in some later event" (italics in original, Bateson, 1971, p. 231;see also, Luhmann, 2013, p. 91). Bateson's very general definition is applied to examine the relation of 'the expected teaching' to the support proposals in IEPs: Teaching is prescribed by curricula and, as mentioned above, contingent on a standard set of adaptation methods. For IEPs to be worthy of their name, they need to contain relevant information beyond these foundations for instructional design and social participation.
Concerning the support proposals, systematic differences are possible. Expectedly, differences in the informational value of these suggestions are related to differences in the competence of the teachers who created them (Al-Shammari & Hornby, 2020;Paccaud & Luder, 2017;Ruble et al., 2010). It is also conceivable that apparent differences in format templates, such as formatting IEPs as lists compared to formatting them as tables, impact the informational value (e. g. Muñoz Martínez & Porter, 2018, p. 10;Ministry of Education Ontario (Ed.), 2004, pp. 59-66). Furthermore, the orientation of IEPs towards specific special educational needs could be visible in the support proposals (Paccaud & Luder, 2017, p. 215). For example, relative to others in a class, pupils with moderate learning difficulties might receive support related to "earlier stages of the learning path" (Fletcher-Campbell, 2005, p. 187). However, connections between the support proposals' informational value and these structural characteristics are largely unexplored, as are contextual features regarding differences between school types, individual schools, and grade levels.

Research questions
(RQ1) What is the proportion of informative support proposals compared to less informative support proposals? (RQ2) What is (a) the proportion of content-specific support proposals per IEP and (b) the extent of their curriculum modification? (RQ3) Which contextual and structural characteristics of IEPs are associated with (a) their proportion of content-specific support proposals and (b) the extent of curriculum modification of these entries?

Data collection and sample description
After obtaining permission from the relevant authorities, IEPs were collected in one federal state in Germany at the beginning of 2019. Participation was entirely voluntary. The sample contained 112 fully anonymised IEPs from seven schools. Although the sample is not representative, it shows a sufficiently broad spectrum of variation.
Concerning the contextual characteristics, coding could be conducted for school type, individual school and grade level. From primary schools came 54.5 % of the IEPs and 45.5 % from secondary schools. Forty-one IEPs (36.6 %) did not indicate grade level. Of those documents whose grade level could be identified, the sample encompassed IEPs from grade levels one to eight. Because the data material was considered confidential, most of it had already been anonymised on-site in the schools during collection. Therefore, coding concerning the teachers who created these documents was not possible in this data collection.
Regarding the structural characteristics, coding was done concerning the format and the explicitly mentioned pedagogical purpose. 60.7 % of the IEPs were written in a table format, 39.3 % in a list format. In most cases (70.5 %), it was unclear whether the IEPs were produced for pupils with identified special educational needs. Of the documents, 24.1 % were related to "special educational needs in learning" (SEN-L; Gorges et al., 2018, p. 12). This category is mainly used in Germany and can internationally best be described as "moderate learning difficulties" (Fletcher-Campbell, 2005;Norwich et al., 2014). Only 5.4 % of the IEPs were explicitly related to other support areas.

Data analysis
Selecting text elements matching the target category "support proposals" was carried out by differentiating the entries in the IEPs according to the document-typical contents mentioned above: present level of performance, goals, support proposals, and entries concerning evaluation. A coding system was developed using "inductive category formation" (Mayring, 2014, p. 79) to identify the types of support proposals. While the statistical analysis unit was the individual IEP document, single propositional units within the IEPs were treated as the coding unit (Flick, 2018;Mayring, 2014). Four main categories of support proposals (see , Table 1) were initially used. The support proposals of types A, B and C appeared to be easily identifiable. Another category, type D, was introduced as a residual category. Its differentiation was postponed, as nearly all propositional units classified as D had little informational value for individualised support.
During two seminars, student teachers carried out the "structuring-deductive category assignment" (Mayring, 2014, p. 95). The student groups' coding was finally compared to a student research assistant who did not participate in either seminar. To assess the reliability of the coding and its practicability concerning the IEPs as the unit of analysis, ICCs were calculated for the various coded characteristics, based on a mean-rating (k = 2), absolute-agreement, two-way mixed-effects model (Koo & Li, 2016). For the frequency of proposals of type A, the ICC absolute (18, 18) was .93. According to the guidelines of Cicchetti (1994, p. 286), this can be judged as excellent. The same applies to the following two types. For the frequency of proposals of type B, the ICC absolute (18, 18) was .91. For the frequency of proposals of type C, the ICC absolute (18, 18) was .88. As expected, since category D was designed as a residual category, it also had low reliability, which must be judged as poor (Cicchetti, 1994): the ICC absolute (18, 18) was .07. As mentioned above, this residual class's entries were subjected to further exploratory, inductive category formation to comprehensively exploit the types and subtypes of support proposals included in the dataset (see , Table 1).
For the subsample of IEPs (n = 47) that included type-A support proposals in German or mathematics and the grade level, these support proposals were assessed concerning their concordance with the respective grade levels' curricula. Analogous to Paccaud and Luder (2017, p. 210), it was rated how many German and mathematics support proposals per document coincide with the respective curriculum's prescribed content. For the number of support proposals classified as grade level accordant per IEP, the ICC absolute (46, 46) was .67. According to Cicchetti (1994, p. 286), this can be judged as a good reliability value. It can also be considered a good reliability value because the phrasing in curricula is often vague (Anderson et al., 2014, p. 4).

Construction of two indicators for the informational value of IEPs
Two different indicators were created to investigate the association of informational value with contextual and structural characteristics of the IEPs. The content-specific support proposals to be carried out in school (type A) were the basis for creating both indicators. This was because the type-A proposals seemed potentially informative for individualised instruction, especially when they contained ideas that deviated from and went beyond the pupils' grade level curriculum. In contrast, the informational value of the other types (B to D) appeared to be redundant for teaching. Above all, it seems necessary to underscore that, while type-B proposals in part named differentiated instruction methods, these entries essentially were mere catchwords: A lack of concretisation was noticeable (see , Table 1). For example, when ability grouping was suggested, the actual teaching method to be employed was not mentioned; similarly, when suggesting tiered assignments, the actual differences in pupils' tasks were left unmentioned. Therefore, these entries' transferability to practice has to be called into question due to the same reason as it is frequently noted in studies on this subject (Deunk et al., 2018;Smale-Jacobse et al., 2019). While type-C suggestions are helpful in coordinating the support provided within and outside school, both indicators generated here aim to assess the informational value of support proposals for in-school support, especially for lesson design. In this respect, type-C suggestions are at best of indirect relevance. Analogous to this, regarding usefulness for lesson Table 1. Types of support proposals in the dataset (N = 1245)

Designation Description and Examples
Type A: 'conventional, contentspecific methods: in school' These were suggestions for teaching and educational methods related to precisely one subject (or one of the domains 'work behaviour' or 'social behaviour') and described how a specific goal could be achieved. Often, concrete teaching material was also mentioned.
'Exercises for segmentation of numbers/quantities; first in the number range up to 5, afterwards up to 10', 'Representation of place values with quadratic, structured fields and additionally with Dienes-Material [certain kind of wooden cubes, small single ones as quantity one, corresponding rods as tens, fields as hundred and big cubes as thousand].' 'Periodic exercises for phonological perception: differentiation and identification of phonemes and letters, e. g. by determining initial, middle and final phonemes; naming of letters and phonemes by announcement and by using phoneme dominoes' Type B: 'conventional but contentunspecific methods: in school' Type B also were methods to be employed in school but without reference to concrete subject content. These practices were not only applicable in a single, specific subject but several. These were mainly organisational proposals or widely known DI methods. planning, type-D entries' overall informational value was also poor, and some of these entries may even have adverse effects.
The first variable for informational value was the Proportion of content-specific support proposals per IEP (Proportion content-specific ). Teachers reading an IEP would have to filter out the support proposals that are rather useless for their lesson planning. Therefore, it makes a difference how many redundant entries such a document contains. Proportion content-specific was calculated as the ratio of content-specific support proposals to all support proposals per IEP. It is the percentage of type-A support proposals per IEP. This ratio was chosen because it allows the desired information to be presented concisely as a percentage value. Beforehand, however, it was checked whether these ratios per IEP sufficiently measure what they were supposed to measure. For this purpose, the Pearson correlation coefficient of Proportion content-specific per IEP with the corresponding less illustrative differences per IEP (number of type-A proposals minus sum of other proposals) was calculated. The variables were found to be strongly correlated, r(110) = .77, p < .001. Therefore, the percentages show what they are supposed to show and are beneficial for reporting due to their general comprehensibility. It was also examined whether Proportion content-specific and the second indicator, described in the following paragraph, measure qualitatively different characteristics regarding the individual documents.
The second variable for estimating the informational value was Curriculum modification. As mentioned above, if a support method deviates in content from what the majority of the class is taught based on the curriculum, this is particularly relevant for lesson planning. Curriculum modification was calculated as the ratio of type-A proposals per IEP that were not curriculumaccordant to the total number of type-A proposals for the respective document. Therefore, Curriculum modification shows the percentage of support proposals for the subjects German and mathematics per IEP, potentially containing deviations from the curriculum for the relative grade level. This indicator was also checked beforehand whether it reflected sufficiently strongly what it was supposed to represent. The Pearson correlation coefficient of Curriculum modification per IEP with the corresponding less illustrative differences per IEP (number of potentially deviating type-A proposals minus curriculum-accordant type-A proposals) was calculated. The variables were found to be strongly correlated, r(45) = .81, p < .001. Again, the percentage value was chosen due to its informational clarity and legibility. As mentioned, it was also examined whether Proportion content-specific and Curriculum modification measure qualitatively different aspects of the support proposals. The Pearson correlation coefficient of these two variables showed a nonsignificant and weak relation, r(45) = .23, p = .125. Against this background, it can be assumed that the two indicators measure different characteristics and represent each of them sufficiently well.

Designation
Description and Examples D 6 These entries mainly were imprecise proposals for promoting critical personality characteristics, e. g.: under the heading 'Cognition': 'Integrate and explain everyday activities', 'Solving riddles'; under the heading 'Perception': 'Paying attention to environment and nature' D 7 Communications of critical issues, especially to parents, which were listed under the rubric of supportive actions e. g.: 'Ensure sufficient sleep', 'Please provide the missing materials by using the material list', 'The school continues to operate as planned. Special needs teacher explains that neither a repetition of the grade level nor removing the ascribed SEN-category is possible.'

Statistical analysis
Descriptive statistics were used to answer the first and second research questions. The third research question was predominantly addressed by using nested analyses of variance (Quinn & Keough, 2002, pp. 208-221) with either Proportion content-specific or Curriculum modification as the dependent variable. The contextual characteristic school type as a fixed factor represents the upper level. The contextual characteristic individual school was a subordinate random factor. Fixed factors that also could be seen as subordinate were the contextual characteristic grade level and the structural characteristics format and explicitly mentioned pedagogical purpose. The nesting of the respective independent variables in the supercategory school type could be applied for most of the analyses. However, if subgroups were lacking due to the nature of the sample (see , Table 4), single factor univariate analyses of variance were conducted. In cases where the assumption of homogeneity of variances was violated, nonparametric tests and according effect size estimations were additionally used as robust methods to ensure the conclusions (Field, 2018, p. 537;Corder & Foreman, 2009;Lenhard & Lenhard, 2016). The ANOVA results are reported throughout for better readability. The level of significance was set to 5 % for all analyses.

Frequency of the different types of support proposals
A total of 1,245 support proposals were identified in the data. Table 2 gives an overview of the types and their frequency of occurrence.

Proportion of content-specific support proposals and curriculum modification per IEP
One IEP contained, on average, about eleven support proposals (M = 11.12, SD = 9.62). Of these, approximately five were content-specific (M = 5.38, SD = 5.07). Thus, in an average IEP, 47 % of the support proposals were content-specific (see , Table 3).
Based on the subsample mentioned above, curriculum modification was calculated using the support proposals in German and mathematics per IEP. Due to the contextual and structural characteristics of these documents (see , Table 4 and Table 5), the subsamples' IEPs showed a higher average number of support proposals in general (M = 13.62, SD = 10.91) as well as of content-specific ones (M = 7.13, SD = 4.94). Most of these proposals concerned the two investigated subjects. For mathematics and German taken together, an IEP contained almost seven support proposals on average (M = 6.60, SD = 3.49). About five were accordant with curriculum specifications for the respective grade level (M = 4.81, SD = 3.51). Thus, the proportion of contentspecific proposals in German and mathematics per IEP in which potential deviations from the curriculum occurred was on average 30 % (Table 3). About one-third of the IEPs (34 %) had no support proposals that deviated from the curriculum. In contrast, a quarter of the IEPs showed potential curriculum modifications in more than half of the content-specific support proposals in these two subjects. Type C: "support methods to take place outside of school" 112 9 Type D: exceptional cases; see subtypes D 1 to D 7 in Table 1 107 9

Association between contextual and structural characteristics of IEPs and their informational value
The descriptive findings for these analyses are shown in Table 4, and the ANOVA results in Table 5.
School type. As shown in Table 4, the proportion of content-specific proposals per IEP was higher in primary schools (52 %) than in secondary schools (41 %). This group difference was significant, with partial η 2 = .04 indicating a small effect (see, Table 5). The differences between primary and secondary schools appeared remarkable when looking at the per-IEP proportion of support proposals with curriculum modification, with an average percentage of 41 % in primary schools and only 11 % in secondary schools. However, due to the sample structure (see , Table 4), the related effect size could also be an overestimation. For interpretation, it should be set aside in favour of the following result.
Grade level. There was a significant effect of grade level nested within school type on the proportion of curricular modified support proposals per IEP, which with partial η 2 = .21 amounts to a large effect. It was further observed that the average proportion of support proposals with curriculum modification per IEP was relatively low in grade level one (17 %), highest in grade-levels three (54 %) and four (53 %), and then clearly lower again in grade-level five (12 %). There was no significant effect of the grade level on the proportion of content-specific support proposals per IEP. Note. According to the available data (see , Table 4), the brackets indicate whether the respective factor was calculated as nested under school type.

Table 5. Results of ANOVAs with the proportion of content-specific support proposals per IEP or the proportion of curriculum modification per IEP as the dependent variable; school type, grade level, format and explicitly mentioned pedagogical purpose (SEN-L) as fixed factors; and individual school as a random factor
Individual school. There were no significant differences between the IEPs of the individual schools concerning their share of curricular modified support proposals but concerning their percentage of content-specific proposals per IEP. The corresponding partial η 2 = .18 indicates a large effect. The differences between IEPs from individual schools were particularly remarkable, superficially using the same format. For example, schools Four and Seven both predominantly used a table format but differed clearly in the proportion of content-specific proposals per IEP (73 % versus 30 %).
A contrasting look at the documents in table format was instructive: At school Four, there were notably fewer columns than at school Seven. These only described the "area of support", the "goals" and the "support proposals". Thus, there was enough space for detailed descriptions available on the page, especially for the support proposals. The rows below the column "support areas" exclusively contained support areas that can be considered typical for school and teaching, mainly German and mathematics as well as work and social behaviour. The specific table format that school Seven predominantly used had six columns: "competence area", "present level of performance", "goals", "support proposals", "who?", and "results". The entries under these rubrics were much shorter due to lack of space on the page and were very repetitive across the respective documents. Pre-made entries seem to have been used frequently. In addition, under the first column, "Areas of competence", numerous subcategories set by the schools' personnel were found, which were titled with psychological terminology such as "Cognition", "Perception" and "Self-Competence", which frequently led to type-D proposals.
Format. The format was also associated with a significant difference in the proportion of content-specific proposals per IEP. The table format showed, on average, higher values (52 %) than the list format (41 %). However, the corresponding significant effect of partial η 2 = .06 was considerably smaller than the aforementioned effect due to the individual schools' differences.
Explicitly mentioned pedagogical purpose. Due to the sample, this factor could only be analysed for the difference between IEPs focusing on pupils with SEN-Learning and those IEPs that did not mention an explicit purpose. The IEPs for pupils with SEN-Learning had a higher average proportion of content-specific proposals (60 %) than those IEPs without an explicitly stated pedagogical purpose (43 %). However, there is an interaction to consider here (see , Table 5). While the proportions of content-specific proposals in IEPs without explicitly mentioned pedagogical purpose were comparable in primary and secondary schools (42 % versus 44 %), these proportions in IEPs with stated SEN-L were increased in primary schools (73 %) and decreased in secondary schools (22 %). Furthermore, IEPs that mentioned SEN-L showed a higher proportion of curriculum modification per IEP (63 % compared to 26 %). Overall, the effects were significant. Concerning the proportion of curriculum modification per IEP, the effect size of partial η 2 = .12 amounts to a medium effect. On the proportion of content-specific proposals per IEP, the effect size of partial η 2 = .15 amounts to a large effect. Also, the mentioned interaction between school type and pedagogical purpose on the proportion of content-specific proposals per IEP was significant and with partial η 2 = .12 amounted to a medium effect size.

Discussion
Basically, in contrast to the results of Özdemir et al. (2020) and Ruble et al. (2010), it can be said that support proposals were found in almost all IEPs. On average, eleven support proposals were included in an IEP. However, entirely in line with Räty et al. (2019), the support proposals analysed in the present study also could often be classified as too vague to enable concrete implementation (see , Table 1 and Table 2). The categories found were not only diverse in the amount of information they provided, but they were also likely to be perceived differently by various actors. As described in the methods section, it can be assumed that the "conventional, content-specific methods: in school" (type A) were the most informative support proposals for all parties involved, especially for school staff and pupils. It was found (see, Table 3) that the proportion of contentspecific support proposals averaged 47 % across all IEPs in the sample. A subsample of the documents showed that per IEP, on average, 30 % of the support proposals for German and mathematics contained deviations from the curriculum for the related grade level. This proportion of variations is slightly higher than Paccaud and Luder (2017, p. 218) found based on their sample.
Specific contextual and structural characteristics were associated with significantly increased informational value. For example, the simple difference between whether these documents were created in a table or list format was associated with the proportion of content-specific proposals per IEP. A significant medium-sized effect indicated that, in tendency, the IEPs in table format had a higher informational value in this respect (see , Table 5). Concerning the proportion of content-specific proposals per IEP, however, there were even more considerable differences between the documents of the individual schools. Moreover, systematic differences occurred even between IEPs of the same format. A contrasting look at schools' varying table format templates with clearly differing average percentages of content-specific proposals per IEP was already revealing, as previously described in the results section. There were no uniform format templates prescribed at the federal state level. Instead, each school seems to have generated its own template. These templates contain different amounts of structural aids and standardised categories and, on a very basic level, provide different amounts of space for elaborating on support proposals. Detailed entries for the content-specific proposals, in particular, could be found in table templates that a) contained fewer columns and b) did not prescribe any categories for support or at least no categories that in tendency went beyond the common instructional in-school practices in an unhelpful, even obstructive way. If no learning and behavioural assessment and support specifications were demanded, the teachers who had to prepare these documents instinctively chose their starting points via the curricula and observations on social and work behaviour. These entries hinted at the teachers' strengths in these areas: Obviously, they were often able and willing to produce concrete proposals. If the space in the cells became insufficient due to an overabundance of columns, frequently only catchwords could be found, often presumably-based on the repeated use of identical phrasing-falling back on the individual school's repository of pre-made support proposals. The templates that by default used non-school categories for the observation and support areas were particularly dubious, such as the above-mentioned "Cognition", "Perception" and "Self-Competence". If valid observations and approved support proposals would have been made according to these areas, e. g. usage of Klauer's training on inductive reasoning (Klauer & Phye, 2008), such recommendations would even be desirable in the IEPs. However, the teachers' use of these categories led to support proposals such as "Solving riddles" or "Paying more attention to nature", which simply cannot be described as holistic. In most cases, they were too undetailed to provide any idea of how to implement them. At the same time, they suggest that there is something to promote regarding critical personal characteristics without having an effective answer. Without a construct-appropriate assessment of skills in these areas and the naming of approved supportive instruments, unfortunately, the teachers have engaged in pseudo-psychology at these points. This phenomenon was not observed in other schools that used templates in which these categories did not appear. Furthermore, regarding the proportion of contentspecific proposals per IEP, a significant effect with a large effect size was found, depending on whether these documents were explicitly created for pupils with SEN-Learning, or no specific pedagogical purpose was mentioned in them. For the IEPs from primary schools, there was an increased proportion of content-specific support proposals in documents with stated SEN-L. Therefore, this result is in line with the findings of Paccaud and Luder (2017, p. 220). However, for the IEPs with stated SEN-L from secondary schools, a decreased proportion of content-specific support proposals was found. This may be due to differences in the organisation of teaching between primary and secondary schools. A child is usually taught by a small number of teachers during primary school, whereas in secondary schools, the teaching is usually done by a larger number of subject teachers. The use of fewer content-specific support proposals in IEPs with stated SEN-L could therefore indicate that a sufficiently detailed overview of subject-specific performances was not compiled, which may have led to more generic support proposals. Furthermore, the proportion of support proposals with curriculum modification per IEP was to the extent of a medium effect significantly higher in the documents for pupils with SEN-Learning than in the documents without an explicitly mentioned pedagogical purpose (see , Table 5). Also, this result does not follow Paccaud and Luder's findings, in whose sample the IEPs of pupils from the cluster "behaviour" contained more curricular modifications (Paccaud & Luder, 2017, p. 219). However, the findings are consistent with expectations. This is because the central diagnostic criterion for determining SEN-Learning is a learning delay of more than one school year, especially in German and mathematics (Gorges et al., 2018, p. 12). Even though behavioural problems may appear simultaneously, the focus with these labels lies on deviations in learning levels compared to the same age group (Norwich et al., 2014). What needs to be explained is the significant association of the IEPs of various grade levels concerning their proportions of support proposals with curriculum modification. This association, calculated under consideration of the nested data structure, amounts with partial η 2 = .21 to a large effect (see , Table 5). In the federal state from which the investigated IEPs originate, the first instance of a grade promotion decision occurs at the end of grade level two, and marks are given from grade level three onwards. These marks are mainly used to decide which subsequent school track should be attended after grade four. The fact that significantly increased values were observed concerning curricular modifications in these grade levels could indicate that the school, in the face of this kind of raised stakes, considers it necessary to document the need for these adaptations in teaching both implicitly, using marks, and explicitly, by the descriptions in the IEPs.

Summary assessment of the examined documents' informational value
The sample contained IEPs with specific characteristics associated with a higher informational value (see , Table 4 and Table 5), especially for teachers and pupils. Higher values were found in IEPs compiled using particular templates in table format, IEPs for pupils with SEN-Learning and also in IEPs from grades two to four. However, the average values from research question two especially indicate that the desired informational value of the documents contained in the sample is too low not only for teachers but pupils and parents as well. Particularly confusing is the high frequency of "conventional, but subject-unspecific proposals" (see, Table 2). On the one hand, the frequency of these entries may indicate that differentiated instruction methods are already widely used, similar to the results of other studies (Labhart et al., 2018;Letzel & Otto, 2019;Pozas et al., 2020). On the other hand, proposals like "differentiated exercises" are meaningless for pupils because they are not the target readership, and for teachers, it may merely be a reminder of routine work still to be done when IEPs are applied to lesson planning (Muñoz Martínez & Porter, 2018;Nilholm, 2020). Ultimately, in the abbreviated, catchword-like manner found in the current data set, it is impossible to judge whether teachers adequately concretise these methods in the classroom. This is particularly regrettable because differentiated instruction methods are considered promising. Given the indications in several studies that concretising the methods of differentiated instruction was done, or at least supported, by researchers, as well as the fact that it is a highly complex task in general which not all teachers feel well prepared for (Smale-Jacobse et al., 2019;Van Geel et al., 2019), a rather sceptical interpretation suggests itself based on the present data. The found catchword-like mentions of differentiated instruction practises (and similar entries also under type D) were abstract to a degree that the precise nature and quantity of the teacher's additional workload was left unclear but was presented in a way that, from the school-external, layperson's perspective of a parent, the extra effort may have seemed significant. Also, the rather organisational entries subsumed under the "conventional but content-unspecific methods: in school", especially the employment of SEN teachers, seemed questionable: In the federal state from which the investigated documents originate, in most cases, these resources are allocated on an institutional level and cannot be re-allocated to exclusively support one particular pupil by an IEP in a binding way (Klemm, 2020;Steinmetz et al., 2021). However, the situation was again presented to the outside reader, i. e. the parents, as if the additional human resources had been exclusively delegated to help their child. In contrast, the mentions of "support proposals to take place outside school" were likely to be informative for parents and were also often explicitly addressed to them. While these entries are to be seen as appropriate for coordination between school and home, they are merely of indirect relevance for lesson planning. Furthermore, the low percentages of deviations from the curriculum among the desirable content-specific support proposals are remarkable (see , Table 3). In general, content-specific support proposals show pupils, but also parents-who are still mostly laypersons regarding the instruction of schoolspecific content-, ways in which they can achieve specific goals. For the teachers, however, most of these proposals aligned with the curriculum and are essentially meaningless because they would have to teach them to the entire class anyway. It is worth keeping in mind that this strictly applies to the type of proposals that were considered the potentially most informative for teachers, meaning the majority of other suggestions are even less meaningful for them, which paints a grim picture of the overall usefulness of IEPs from an instructional and pedagogical perspective. Based on the current findings, it becomes apparent why teachers regularly consider the effort required to produce IEPs as too high compared to their benefits (Hirsh, 2014;Moser Opitz et al., 2019;Müller et al., 2017).

On the meaning and function of the IEPs examined
The support proposals tended to contain what teachers would or at least should do (Everitt, 2017), even without an IEP. From their perspective, these entries were mainly repetitions of known information and, thus, had essentially no informational value in Bateson's strict definition (Bateson, 1971, p. 231;Luhmann, 2013, pp. 216-217). Nevertheless, each document consists of at least one, but often two or more pages of text, including, among other things, eleven support proposals on average. Most parents and pupils are not familiar with the curricula nor with common teaching methods. Typically, they are also not aware of how staff resources are allocated. To interpret their perception of "IEPs", it is essential to consider that the document is called "individual" educational plan. This point is probably especially significant for parents. While pupils can still observe in class that others may also be provided with support, perhaps even the same types of support, parents lack this critical basis for comparison. As parents tend to be unable to assess the extent to which proposals listed in an IEP deviate from the standard curriculum and support methods, they might easily believe that everything in the IEP is individually tailored to their child. Therefore, it might look like a great amount of special effort and extraordinary ideas solely for their child. The analyses have shown that this is not the case for most IEPs from the present sample. Even though many teachers legitimately document how well they teach diverse learners and make efforts in the classroom in general, the purpose of IEPs seems to be mainly to portray these efforts as entirely based on individual needs. Due to the documents always being written for a single person, all listed actions implicitly appear to be efforts for a particular pupil. Given these ostensible efforts, should this pupil continue to exhibit bad behaviour or no improvement in performance, the school is no longer so easily made responsible. The burden of proof and progress is shifted to other parties, such as the parents or the pupils themselves (Andreasson & Carlsson, 2013). While this insight, seen globally, is not new (Isaksson et al., 2007;Mitchell et al., 2010), the present study allows a detailed look into the mechanism of how the IEPs from the current sample work. By essentially occupying parents' perception with proposals only impressive to the layperson, especially the implicit portrayal of standard practices as tailor-made for their child exclusively, IEPs, at least most of the ones examined here, serve to immunise teaching from parental criticism.

Practical implications
Based on the present results, making IEPs beneficial for the distinct involved parties seems necessary but also possible. Since the proportion of "conventional, but subject-unspecific proposals" was substantial (see , Table 2), this sort of entries is chosen here as a starting point for recommendations to improve the support proposals section in IEPs. Two fundamental recommendations arise concerning this type of entries. The first is to avoid mentions of didactic and organisational measures that are not implemented solely because of the pupil with an IEP. If it is, nevertheless, considered necessary to document (for the world outside the schools' walls) a supportive intervention not solely aimed at one individual, then it should at least be made clear that it is implemented for the whole class or a subgroup. The second recommendation, which mainly concerns the type-B proposals, is to concretise them. As is known from studies on the effectiveness of differentiated instruction methods, it makes a crucial difference whether only the surface of the measure, such as ability grouping, was mentioned or whether conducive content or teaching strategies were also specified, through which the differentiation method achieves its purpose (Smale-Jacobse et al., 2019, p. 4; Deunk et al., 2018, p. 42). It seems necessary to briefly describe a supportive activity's organisational, content-related and didactic aspects instead of noting catchwords such as "Adaptation of tasks to the learning level". Such an entry could easily be concretised, e. g.: "Tasks adapted to the learning level for crossing the tens boundary are presented using software X and textbook Y". Instead of "work in small groups", one could concretise "reading comprehension exercises in small groups to enable reciprocal teaching strategies, starting with summarising and predicting", if appropriate.
In these examples, as in positively evaluated intervention studies, it is immediately apparent that promising interventions reference content or are entirely based on content. For teachers, the knowledge of a set of instructional methods is also essential (Parsons et al., 2018;Van Geel et al., 2019), but precisely because the IEPs are also supposed to be informative for teacher assistants as well as the pupils, parents and third parties, a behavioural description or subject content establishes not only the conceptual working basis of the interventions but also the basis for comprehensibility. Against this background and the content-specific support proposals (type A) found in the data material, it seems recommendable to increase their proportion in the IEPs. These entries should be described in enough detail, ideally referencing the specific didactical materials, so that parents, teacher assistants, and especially pupils can gain an in-depth understanding of how they can achieve a set goal. Thereby, it should be marked-e. g. by using an asterisk after an entrywhether deviations from the curriculum or standard teaching methods are employed (Goepel, 2009, p. 126;Muñoz Martínez & Porter, 2018, p. 10). The same applies to the "support proposals to take place outside school" (type C). For these, too, it would be desirable to give, for example, parents and educators in the daycare setting some advice on instructional strategies and which didactic materials they should practise with. Apart from this recommendation, these support proposals, serving the coordination between school and home, are meaningful and, therefore, should be maintained in IEPs.
In contrast, entries of subtypes D 4 to D 6 -which emphasise "the special educational" aspect or engage in pseudo-psychology without delivering effective supportive instruments-should be avoided. With its various subcategories (see , Table 1), residual category D showed, above all, that imprecise and implicitly pejorative descriptions of interventions could appear in practice. The found negative patterns can be used to avoid such errors consciously.
Contrasting the format of the templates used by the individual schools showed that observational categories such as "Cognition", "Self-Competence", "Perception" and the like should be avoided, unless the involved personnel has the necessary qualification to assess these areas and implement measures adequately. A high informational value seemed to be generated by templates in which classical school domains (subjects as well as work and social behaviour) were elaborated on.
In general, the table format showed two advantages. Firstly, the descriptions of the present level of performance, learning goals and support proposals were more clearly related than in the list format. This matches recommendations to base goals and intervention designs on insights from curriculum-based assessments (Xu & Kuti, 2021). As a consequence of the first advantage, a second one emerged. Due to clearer allocation, most observations and learning goals apparently were "answered" with a support proposal, whereas, in the list format, the associations were less systematic and, thus, the proposals' completeness was more challenging to grasp.
In sum, to generate adequate support proposals, first, they need to be based on and clearly linked with learning goals derived from descriptions of the present level of performance. Second, the support proposals should be more than a catchword. It is recommendable to formulate both the content or skill to be learned and the form in which this could be achieved. For instance, contents could be grouped under the various formats of didactic and organisational arrangements mentioned in the investigated IEPs (see examples of type-B and -C proposals in Table 1). By complementing them with a set of content-e. g., under "Weekly plan", "Educational games", "Help given by the teacher" and concerning the learning arrangements outside of school-concrete and, thus, promising proposals would emerge. Third, if there are deviations from standard teaching methods or the grade level's curriculum, these deviations should be made visible, as mentioned above. Vice versa, non-marked proposals would indicate the respective pupil's capability to keep up with the curricular requirements or to participate in standard instructional situations. Such clarity could make support proposals in IEPs less disheartening and instead offer a more helpful compilation of hints to pupils and parents as well as to support staff; at the very least, it would provide a basis for constructive discussion between all involved.

Limitations
The sample collected was not representative and showed some missing data. For example, a notable amount of information was missing from the IEPs on grade levels and special educational needs. While the results of the analyses can be considered well-founded, attempts should nonetheless be made in future data collections to fill in gaps in these documents through explicit supplementary requests. It would also be desirable to have more resources for analysing the data, including the subcategories from the former residual category. Furthermore, it would be desirable to collect data on the teachers and other professional staff involved in preparing the IEPs. Finally, the pupils' perceptions of these documents should be investigated further (Faldet & Nes, 2021;Goepel, 2009). In the few studies that measured the effectiveness of IEPs on learning outcomes by means other than surveys, neither the content of the IEPs was addressed nor were references made to the classmates' results (Barnard-Brak & Lechtenberger, 2010;Sinclair & Poteat, 2018). A design in which the effectiveness of qualitatively different IEPs could be determined utilising standardised pre-and post-tests in central achievement domains and concerning behavioural dimensions would be illuminating.