phi.intra: Intra-group Phi

Description Usage Arguments Details Value Author(s) References See Also Examples

Description

Calculates intra-group phi for a group of participants, each of whom has free sorted the same set of items. Intra-group phi is a measure of agreement between free sorters within a group.

Usage

1
phi.intra(edta)

Arguments

edta

A data frame that includes the following columns: SjNo, CategoryNo, ItemNo, described below. Each row is one classification decision.

Details

Dataset lawson provides an example of the format expected by the phi.intra function. Only the following columns are used, other columns will be ignored:

SjNo - integer; must be different for each participant.

CategoryNo - integer. For each participant, it should range from 1 to the number of groups that participant produced. Each group that participant produced must have a unique number. Other than that, the numbers are arbitrary.

ItemNo - integer; ranges from 1 to the number of items in the stimulus set. Item numbering must be consistent across participants, e.g. if ItemNo 21 is 'duck' for participant 201, it must also be 'duck' for all other participants in the data frame.

Each row of the data frame is one classification decision. For example, one row might indicate that subject 204 placed item 21 into category 3. The list must be complete for every participant. For example, if the stimulus set contains 20 items then there must be exactly 20 entries for each participant. If a participant fails to classify an item, include the item number in a row but set the CategoryNo to 'NA'.

Cramer's Phi (Cramer, 1946) is calculated for each pair of participants, and the mean of these values is returned. Phi-intra ranges between 0 and 1; 1 indicating perfect consistency among sorters (within the definition of consistency employed; see Lawson et al., 2017, appendix for details). The level of agreement expected among randomly responding classifiers can be calculated using phi.intra.chance.

The phi-intra metric is based on Cramer (1946), slightly developed by Wills & McLaren (1998), and first employed in its current form by Haslam et al. (2007). The Appendix of Lawson et al. (2017) provides a brief tutorial.

Value

Phi-intra; a number ranging between 0 and 1.

Author(s)

Andy J. Wills (andy@willslab.co.uk)

References

Cramer, H. (1946). Mathematical models of statistics. Princeton, NJ: Princeton University Press.

Haslam, C., Wills, A.J., Haslam, S.A., Kay, J., Baron, R. & McNab, F. (2007). Does maintenance of colour categories rely on language? Evidence to the contrary from a case of semantic dementia. Brain and Language, 103, 251-263.

Lawson, R., Chang, F. & Wills, A.J. (2017). Free classification of large sets of everyday objects is more thematic than taxonomic. Acta Psychologica, 172, 26-40.

Wills, A.J. & McLaren, I.P.L. (1998). Perceptual learning and free classification. Quarterly Journal of Experimental Psychology, 51B, 235-270.

See Also

phi.inter, phi.intra.chance

Examples

1
2
3
  data(lawson)
  edta <- lawson[lawson$ExptGroup == 2,]
  phi.intra(edta)

freesortphi documentation built on May 2, 2019, 5:54 p.m.