TABLE 1.

Overview of analysis techniques used to study the role of neural oscillations in speech and music processing, the possible inferences that can be drawn when utilizing each technique, and the references in the present mini-review that applied this technique in older adults, clinical populations or music processing (see in-text references for additional context regarding each study).

Technique	Description	Inference drawn from this technique	Studies cited in this mini-review applying this technique	Key findings
Cross-correlation (Figure 1A)	- Correlation between time series of neural oscillations and lagged time series of stimulus features (envelope, periodicity) is assessed to obtain cross-correlation function	- Fidelity of neural response is encoding stimulus features - Latency of neural tracking	Mirkovic et al. (2019), Braiman et al. (2018)	Mirkovic: Hearing aid simulator employing a directional microphone led to faster neural processing of speech envelope Braiman: Fidelity of cortical tracking of speech envelope in individuals with severe brain injury who successfully performed fMRI mental imagery task is comparable to controls
Multivariate temporal response functions (TRFs) (Figure 1C)	- Regression between time-lagged (to account for neural latency) stimulus features (envelope, phoneme onsets, semantic dissimilarity, etc.) and neural oscillations to predict a temporal response function (TRF) model that explains the mapping between stimulus and neural oscillations - Can be used to reconstruct the stimulus envelope from neural responses	- Time course and source of neural regions tracking stimulus features - Fidelity of representation of stimulus features through model fit - Cannot directly infer entrainment but conveys information about tracking	Di Liberto et al. (2018), Dial et al. (2021), Decruy et al. (2019, 2020), Brodbeck et al. (2018), McHaney et al. (2021), Gillis et al. (2022)	Di Liberto: Atypical cortical tracking of phonetic features in children with dyslexia, particularly in the right hemisphere. Magnitude of cortical tracking correlated with phonological processing abilities. Dial: Increased cortical tracking of speech envelope at early and late latencies in individuals with logopenic variant primary progressive aphasia vs. age- and hearing-matched controls in theta, but not delta, band Decruy et al. (2019): Supralinear increase in cortical tracking of speech envelope for speech in noise in older adults vs. younger adults. Cortical tracking associated with speech comprehension. Decruy et al. (2020): Larger increase in cortical tracking of speech envelope for attended vs. unattended speech in individuals with hearing loss vs. age-matched controls. Brodbeck: Increased cortical tracking of speech envelope in older adults reflects inefficient recruitment of regions outside of primary auditory cortex at early latencies. McHaney: Better speech-in-noise comprehension was observed in older adults in whom competing noise showed less deleterious effects on delta band tracking of speech envelope. Gillis: Increased cortical tracking of speech envelope and delayed latencies in individuals with hearing loss vs. age-matched controls. Age-matched controls, but not individuals with hearing loss, showed increasingly delayed latencies with greater background noise.
Mutual information	- Assesses statistical dependency between bandpassed stimulus rhythms and neural oscillations - Analysis typically performed at multiple time lags and averaged across time lags - Can also be used to infer statistical dependency between disparate stimulus and oscillatory bands	- Amount of information that is shared between stimulus and neural oscillations in spectral or temporal domains	Zan et al. (2019)	Zan: Reduced mutual information between neural responses and stimulus with greater noise in older adults. Older adults show greater reduction in mutual information when competing signals are changed from meaningless to a meaningful speech, suggesting age-related informational loss.
Inter-trial phase coherence (Figure 1B)	- Coherence between the phases of each frequency in every trial estimated while ignoring their absolute magnitude	- Phase locking of neural responses and consistency in phase alignment	Doelling and Poeppel (2015), Sorati and Behne (2019), Yu et al. (2018)	Doelling and Poeppel: Musicians showed enhanced cerebro-acoustic coherence across a range of tempi while nonmusicians demonstrated similar coherence only at 1/sec and higher. Degree of coherence correlated with ability to detect pitch distortions. Sorati and Behne: Lower inter-trial phase coherence for musicians and non-musicians in delta, theta, and beta bands in audiovisual speech perception. Desynchronization in alpha band for audiovisual speech only in musicians. Yu: Inter-trial phase coherence to speech in individuals with autism spectrum disorder increased at earlier latencies and decreased at later latencies vs. controls.
Cerebro-acoustic coherence	- Coherence between stimulus envelope and neural activity obtained using cross-spectral density estimates - Procedure focuses on how individual frequency components of neural oscillations relate to individual frequency components in stimulus envelope - Cannot be used with discrete stimulus features such as word onsets	- Phase-locking of the envelope frequencies and M/EEG spectral components - Informs entrainment in restricted sense but is difficult to separate from evoked activity	Harding et al. (2019), Vanden Bosch der Nederlanden et al. (2020), Mandke et al. (2022), Fiveash et al. (2020), Molinaro et al. (2016)	Harding: Cerebro-acoustic coherence to music rhythm increased with years of musical training while response to speech rhythm did not differ as a function of musical training. Vanden Bosch der Nederlanden: Under easy listening conditions, neural phase-locking is comparable for spoken sentences vs. sung sentences, but under challenging conditions, better neural phase-locking observed for sung speech, particularly in the theta range Mandke: Decreased neural coherence to speech envelope in children with dyslexia in 0-5 Hz and 12-40 Hz range. Fiveash: Adults with and without developmental dyslexia showed enhanced stimulus-brain coherence for regular vs. irregular rhythms in music, but individuals with dyslexia did not extract subtle temporal regularities from irregular stimuli. Suggests top-down contributions to neural processing of music. Molinaro: Individuals with dyslexia showed impaired entrainment to speech and reduced stimulus-brain synchronization in delta band in primary auditory regions relative to controls.
Cross-frequency coupling (Figure 1D)	- Degree of phase-to-phase or phase-to-power alignment between two different oscillatory frequency bands - Estimated by obtaining the instantaneous phase of a low frequency oscillation and assessing its phase coherence with the instantaneous amplitude envelope of a higher frequency oscillations	- Interaction between oscillations in different bands - Relationships across perceptual timescales or causal relationships between top-down and bottom-up processing	Power et al. (2016)	Power: Children with dyslexia showed significantly poorer speech encoding in 0–2 Hz band compared to both chronological and reading age-matched controls. No group differences were found between delta phase and beta power coupling suggesting no differences in sensory-motor coupling between individuals with dyslexia and controls.