The left-right self-placement can measure generalized political positions and ideological orientations. It is based on the usage of the words "left" and "right" for different political attitudes. This language convention is derived from the seating order of the French National Assembly in 1789 (Converse and Pierce, 1986). To date, the terms "left" and "right" belong to every day's language usage and reflect people's stands on major political objects. In Germany a right political orientation suggests a rather hierarchical, conservative and nationalistic attitude, whereas a left orientation suggests rather a socialistic, progressive and international orientation. Between these two extremes, there are various gradations. When political parties can be clearly assigned to left or right, the left-right self-placement is correlated with party choice (Inglehart & Klingemann, 1976). It reflects rather an ideological self-identification than a value orientation (Knutsen, 1998).

The advantage of the left-right self-placement lies in its simplicity. Even complex political circumstances can be displayed in a comparably easy way and thereby be classified in the overall political context (Fuchs & Kühnel, 1990). The scale is often used for self-placement and external assessment of political attitudes and to position political parties on the left-right spectrum.

Item generation and selection

The scale has already been applied in the 1960s with similar wording and various response specifications. Deutsch, Lindon and Weill (1966), for example, used a seven-point scale from left to right to assess political attitudes in France. In ALLBUS, the scale is used since 1980 in every ALLBUS round (except 1984). The item and the number of answer categories remained unchanged (1 = left, 10 = right). Using a 10-point scale instead of an 11-point scale has the advantage of avoiding a tendency towards the center because of low motivation or missing knowledge about the meanings of "left" and "right" (cf. Scholz & Zuell, 2012). The labels of the categories changed over years. From 1980 to 1992, only the very left and the very right categories were labelled with "left" and "right". From 1993 to 1998, all categories were additionally numbered from 1 to 10. Since 2000, categories are labelled with letters (F=1, A=2, M=3, O= 4, G= 5, Z=6, E=7, Y=8, I=9, P=10). Naming the categories with letters should prevent possible positive or negative connotations due to numbers or words in the answer format. Additionally, there is the answer category "not specified".

Samples

Sample 1 is derived from ALLBUS/ISSP 2014 (Left-Right Self-Placement with 10-point rating scale). The population consists of all people over 18 years living in private households in Germany (Germans and foreigners). Out of this population, a two stage disproportionate random sample in Western Germany (incl. West Berlin) and Eastern Germany (incl. East Berlin) was drawn. In the first sample stage municipalities in Western Germany and in Eastern Germany were selected with a probability proportional to their number of adult residents; in the second sample stage individual persons were selected at random from the municipal registers of residents.

Targeted individuals who did not have adequate knowledge of German to conduct the interview were treated as systematic unit non-responses. The sample has been weighted by population such that 2,400 interviews are conducted in the Western states and 1,100 in the Eastern states. Therefore, design weights (east-west-weights) were used for the following analyses, to obtain representative results for Germany. The final ALLBUS/ISSP sample consists of n = 3,471 persons, thereof n = 1,763 male and n = 1,708 female with an average age of 49.0 years (sample 1).

For reliability and validity estimation, additional samples were used. Sample 2 is derived from the German Longitudinal Election Study (GLES) 2005-2009-2013 (Left-Right Self-Placement with 11-point rating scale). Participants were multi-level randomly selected out of all persons living in private households in Germany who were eligible to vote in 2005 and who had a minimum age of 18 years. The sample was stratified via federal states, administrative regions and community types. In 2005, the total sample 2 consisted of n = 2,540 persons, thereof n = 1,257 male and n = 1,283 female with an average age of 50.3 years.

Sample 3 is derived from the GESIS Panel 2015 (Left-Right Self-Placement with 7-point rating scale). Participants were multi-level randomly selected from the registration offices. In 2013, the final sample consisted of n = 7,599 persons, thereof n = 3,746 male and n = 3,853 female with an average age of 59.2 years.

Item parameter

Table 2

Means and standard deviations of the left-right self-placement (ALLBUS 2014)

Note. Observations were weighted based on east-west weights. Scale from 1 ("left") to 10 ("right"),

n = 3,263.

Objectivity

Data from ALLBUS/ISSP 2014 were collected by trained interviewers in personal interviews with standardized questionnaires (CAPI – Computer Assisted Personal Interviewing) and two additional self-completion questionnaires (CASI – Computer Assisted Self-Interviewing) for ISSP. Therefore, objectivity of application is ensured.  As the scale is objectively rated by using fixed categories and a norming sample (see descriptive statistics), a high objectivity of interpretation is achieved.

Reliability

As the scale consists of one item only, the reliability estimation is based on longitudinal data. The data from different German long-term studies were correlated to compute retest reliability. First, ALLBUS 2014 data were correlated with data from the ISSP 2014. The ISSP questionnaire was completed directly after the ALLBUS questionnaire. The degree of correspondence of different test results from the same participants at both times is r_tt = .86 (p < .001, n = 1,575).

In the German Longitudinal Election Study (GLES), the respondents completed the items every two years (2005 – 2013). Here, retest correlation between two successive years ranged from r_tt = .27 (years 2009 and 2011, p < .001, n = 430) to r_tt = .60. (years 2011 and 2013, p < .001, n = 409). In the GESIS Panel, also with a retest interval of two years (2013 – 2015), retest correlation was still r_tt = .42. (p < .001, n = 7,599).

The data of the ALLBUS/ISSP 2014 suggest that the item allows a reliable measurement. On the other hand, the data of the GLES and the GESIS Panel suggest that left-right orientation is only moderately stable over time. This result is consistent with observations of typical changes in the political attitude of individuals over time, according to their view of political parties or current political situations.

Validity

To ensure content validity, the construct should be defined in advance and all aspects of the construct should be considered. As the left-right self-placement is based on the common language conventions of “left” and “right” which are well-known and defined in the literature, content validity can be seen as given.

Criterion validity is given if the scale measures the constructs in a way that it is consistent with an external criterion. For this purpose, the left-right self-placement can be correlated with other criteria from ALLBUS/ISSP 2014. A higher scale value stands for a politically more right attitude. According to Inglehart and Klingemann (1976), self-placement is correlated with party choice, when parties can clearly be classified as left or right. A high left-right self-placement score is expected to correlate positively with a preference for parties typically classified as right-wing. Conversely, there should be a negative correlation with parties typically classified as left-wing. Left-right party-placement in the German GLES study shows that a big majority of people sees the CDU party as rather right-oriented, whereas the party Die Linke is considered to be typically left-oriented. Therefore, preferences of these parties were correlated with left-right self-placement in the GLES study.

Considering the construct of right-wing authoritarianism (Altemeyer, 1998), a right-wing political attitude includes several forms of authoritarian features like authoritarian submission and conventionalism. In ALLBUS 2012, in addition to the left-right self-placement, an authoritarianism scale was included. Therefore, both scales were correlated.

Furthermore, research has shown that there are associations between Big-Five personality and left-right self-placement (see Sibley, Osborne, & Duckitt, 2012 for a meta-analysis,). Openness is often found to predict liberal ideology, whereas conscientiousness predicts conservative ideology. These correlations were tested with the Big Five Inventory (BFI-10; Rammstedt, B., Kemper, C., Klein, M.C., Beierlein, C., & Kovaleva, A., 2012) within the German GESIS Panel data (2013).

The following items were used to test criterion validity of the left-right self-placement:

-       Party preference (GLES 2013): "Was halten Sie so ganz allgemein von den einzelnen politischen Parteien?“ (“What do you think in general about the different political parties?”)

-       Authoritarianism (ALLBUS 2012; α = .56):

-       “Wir sollten dankbar sein für führende Köpfe, die uns genau sagen können, was wir tun sollen und wie.“ („We should be thankful for leading figures who tell us what to do and how.“)

-       “Im Allgemeinen ist es einem Kind im späteren Leben nützlich, wenn es gezwungen wird, sich den Vorstellungen seiner Eltern anzupassen.“ („In general, it is useful for a child in future life if it is forced to adapt to the ideas of its parents.”)

-       Openness (GESIS Panel 2013, α = .95):

-       „Ich habe eine aktive Vorstellungskraft, bin fantasievoll.” („I see myself as someone who has an active imagination.”)

-       “Ich habe nur wenig künstlerisches Interesse.“ („I see myself as someone who has few artistic interests.”) (-)

-       Conscientiousness (GESIS Panel 2013, α = .90):

-       “Ich erledige Aufgaben gründlich.” (“I see myself as someone who does a thorough job.”)

-       “Ich bin bequem, neige zur Faulheit.“ („I see myself as someone who tends to be lazy.”) (-)

Table 3

Correlations of the left-right self-placement with relevant variables

Note. Pearson correlation coefficients, Party preference (GLES 2013): n = 479, Authoritarianism (ALLBUS 2012): n = 3,317, Openness/Conscientiousness (GESIS Panel 2013): n = 7,599, ** p < .01 (two-tailed).

Practical importance of the validity coefficients can be interpreted according to Cohen’s (1992) standards: small effect (r = .10), moderate effect (r = .30), strong effect (r = .50).

Consistent to expectations, there is a strong association between a right political attitude and the preference for the German right-oriented party CDU and a medium negative correlation with preference for the German left-oriented party Die Linke. There is also a small positive effect of the correlation between a right political attitude and Authoritarianism. In the same vein, there is a small negative correlation with openness.^[1]  However, for conscientiousness, no association to left-right self-placement was found. All other expected associations could be confirmed by data. Altogether, the results indicate construct validity for the left-right self-placement.

Descriptive statistics (scaling)

The value for skewness is .05 and the value for kurtosis is .29.  According to Miles and Shevlin (2001), deviations from normality ≤ 1 can be neglected. As both the values for skewness and kurtosis are ≤ 1, the assumption of a normal distribution can be assumed to be given. This is also confirmed by a comparison with the normal probability plot.

In Table 4, reference values in terms of means and standard deviations for the left-right self-placement scale are presented. Data are retrieved from the German General Social Survey ALLBUS 2014. They were weighted based on east-west weights to obtain population representative values. The split of the educational categories is based on the kind of school leaving certificate. Low education means no certificate or lowest level (“Volks-/Hauptschule”), medium education indicates intermediary level (“Mittlere Reife”) and high education includes qualifications for universities of applied sciences (“Fachhochschulreife”) and other universities.

Table 4

Reference values for the left-right self-placement

 Note. Observations were weighted based on east-west weights, n = 3,254.

Further quality criteria

For test economy, there should be a high knowledge profit with, at the same time, little effort in terms of financial and time costs (Moosbrugger & Kelava, 2012). Compared to its shortness and little implementation time, the one-item left-right self-placement scale provides a big amount of new knowledge about the participant’s political attitude. The scale also provides more meaningful and enduring information than the identification with some specific political parties (Converse & Pierce, 1986). Equally, preparation and evaluation take little time, especially in computer-supported studies with automatic data recording, but also when the scale is asked orally, face-to-face or in paper-pencil format. Finally, the use of the scale is not associated with financial costs as it is available for free. Therefore, test economy is a major benefit of the left-right self-placement.